<\/p>\n Second revised edition. Original in Swedish. This report gives an introduction to the subject of artificial intelligence (AI) and suggestions for how the healthcare sector can begin to explore and understand it.<\/p>\n The authors have studied the options for offering automated decision-making support both for those working in healthcare and for patients in their daily lives.<\/p>\n Three different hypotheses based on human senses were examined:<\/p>\n \u201cWe tried to approach these hypotheses in a modest way so as not to reject them, but more to consider how they can support patients and healthcare professionals in their work to improve the content and availability of healthcare, focussing on people-centred healthcare.\u201d<\/p>\n The book is the result of a preliminary study on AI and machine learning, supported by Region V\u00e4stra G\u00f6taland\u2019s innovation fund.<\/p>\n The revised edition was developed with support from the strategic innovation programme Swelife.<\/p>\n Artificial Intelligence (AI) and machine learning are on everyone\u2019s lips, not least after the focus it was given by the Swedish physicist Max Tegmark in book form and on TV. We often associate artificial intelligence with super-smart computers that win at chess and Jeopardy or with a changed job market where many tasks now \u2013 or in the near future \u2013 can be carried out with the help of digitisation.<\/p>\n But does artificial intelligence have an impact on something as tangible and human as healthcare? Of course. It is a prerequisite for the future of health and medical care. AI is a success factor, not least in prevention and early detection. By using health data strategically and systematically, we can make healthcare more efficient and hopefully also cheaper. Public health can be improved.<\/p>\n The amount of data on each individual\u2019s health status is growing rapidly. This includes the information the patient gives during visits within the healthcare system \u2013 such as all the data that a standard blood sample provides \u2013 as well as data that the individual produces, for example using health apps. All data on the individual that is of importance to health and is collected in a structured and systematic way is known as systematic health data.<\/p>\n In the right hands, the available health data could be used to provide health and medical care that is customised to each and every one of us. Unfortunately, the work of the medical care system and regulatory authorities is lagging behind, which means we cannot use our health data to the full. This is something Swelife is working on through the SWEPER project, where among other things we look at the legal, regulatory and semantic obstacles on the path to systematic health data, which leads to better use of the data \u2013 and in the longer term to better public health and a strong life science sector in Sweden.<\/p>\n It is now a matter of urgency. Sweden could have a competitive advantage if we quickly straighten out the issues surrounding AI, machine learning and systematic health data. We have great advantages compared to many other markets, such as social security numbers and a health and medical care system available to all.<\/p>\n To get full benefit from AI investments in the future, we must invest in structural changes. The changes are perhaps not as stimulating to the imagination as chess-playing computers, but they are necessary in order to really be able to harness the power of artificial intelligence and machine learning.<\/p>\n We must also think globally from the start and create an international connection, where we work in accordance with international standards, code systems and regulations so that we can share data across borders.<\/p>\n It is important that Sweden does not just become a supplier of health data, but that we build up and retain the competence to process data. This will create value for us as individuals and for society as a whole. Expertise such as this could become a new export product.<\/p>\n If Sweden is to continue to be a world-leading life science nation the entire life science sector must have better access to systematic health data. Researchers from both industry and academia must be able to share the data. Otherwise we will fall behind in terms of the development of new diagnostics, medical technology, medicines and new treatment methods.<\/p>\n Marcus \u00d6sterberg and Lars Lindsk\u00f6ld\u2019s preliminary study was carried out with the support of Region V\u00e4stra G\u00f6taland. The preliminary study is an excellent example which illustrates the challenges we face. Swelife has now published a somewhat revised edition so that the book can be widely distributed.<\/p>\n Lund, 22 October 2018 Our intention has been to investigate the suppliers\u2019 offer as well as to find out how to proceed with the basic groundwork itself.<\/p>\n At this point, we must state that it is not artificial intelligence (AI)<\/a>1<\/a> we worked with but rather machine learning (ML)<\/a>2<\/a>. The people who seem to take a realistic view of the hype surrounding AI seem to agree that machine learning is a sub-division within AI in academic circles, but that we will have to wait at least a few decades before we have meaningful AI. So in the report we can perhaps relate to the hopes for AI in the future, but we want to be clear that we believe that machine learning is a much more appropriate term for the state of technology development today. Possibly even that machine intelligence, which is used a lot in academic contexts, sets the right expectations.<\/p>\n First, we need to distinguish between the concepts of medical history<\/a>3<\/a> and patient accounts. In this report, medical history refers to the disease history recorded by health professionals. It is admittedly supplied by the patient at the time of care, but it is a controlled conversation with the aim of getting a good overall picture, as those of us in the healthcare system want to structure the history in order to provide the background to what we are doing.<\/p>\n A patient account, on the other hand, is told more spontaneously and in other contexts.<\/p>\n The reason that we need to differentiate is that in the healthcare system we have a lot of medical history connected to visits. Accounts may be recorded in an app in the form of a health diary, or in any uncontrolled way in which patients log their health. Medical history can be described as information given at a certain time and the individual\u2019s account as information which covers a period of time.<\/p>\n Using NLP (Natural Language Processing)<\/a>4<\/a>, we can find out what a person is talking about, and what symptoms they have (using NER, Named Entity Recognition<\/a>5<\/a>). In this way we can look up medical code systems, care plans and guidelines to find an appropriate activity. During the preliminary study, we have mainly matched medical histories of the thorax sector with the code system ICPC (International Classification of Primary Care)<\/a>6<\/a>. However, the method can of course be used with other common code systems, e.g. Snomed CT<\/a>7<\/a>, ICD-10<\/a>8<\/a> and KV\u00c5<\/a>9<\/a>.<\/p>\n When we evaluated Amazon\u2019s AWS service for NLP, we are informed in a friendly but firm way that it does not support Swedish, and we are not confident that automated translation will not lead to the loss or distortion of information.<\/p>\n However, it seems to be possible to make NLP more manual (and in Swedish) through frameworks such as NLTK (Natural Language Toolkit)<\/a>10<\/a> and you can process what you obtained with other machine learning techniques, for example deep learning<\/a>11<\/a>.<\/p>\n A major challenge is the lack of structured information about diagnoses, medical guidelines etc. which can be understood by a computer. If this type of information was machine-readable instead of in the form of PDF files, and computers could investigate it as human beings can investigate Wikipedia, we would be able to achieve much more. This is done using a technique called linked data<\/a>12<\/a>. The Swedish National Board of Health and Welfare with its API \u201cF\u00f6rs\u00e4kringsmedicinskt beslutsst\u00f6d\u201d (Medical insurance decision-making support) has understood this.<\/a>13<\/a><\/p>\n A promising thesis was presented at Chalmers University of Technology in the spring about summarising medical texts. We have been informed by psychiatric services in Region V\u00e4stra G\u00f6taland that this would be an attractive solution since their patient records tend to be very long and difficult to read properly prior to each visit.<\/p>\n Relying solely on a Swedish voice-based conversation as an interface does not currently seem possible. Often misunderstandings arise and the question is whether we gain anything if the user must manually correct what has been said. We have mainly tested Apple SiriKit and Microsoft Azure. Azure did not understand what we said even when we chose to read a certain line from The Godfather films, not even after three attempts, despite the fact that it knew which line it was. Admittedly the test was performed in a noisy environment, but it is certainly not an unreasonable sound image for a realistic user scenario, in our opinion.<\/p>\n We have also evaluated the smart speakers Google Home and Amazon Alexa in order to consider what equipment a user would be able to have at home when they need to contact the healthcare system in future. These two speakers do not speak Swedish, but for those Swedes who speak one of the world languages \u2013 such as Spanish, Cantonese or Arabic \u2013 similar gadgets may become relevant faster than for the majority of Swedes.<\/p>\n What we have learned from testing voice-controlled gadgets is that they are great for those who, for one reason or another, have difficulties with writing, spelling or reading, but can speak just fine.<\/p>\n ”A common assessment is that 5 to 8 per cent of the population in the literate part of the world has problems with reading and writing of a dyslexic nature.\u201d<\/p>\n \u2013 The Swedish Dyslexia Association<\/a>14<\/a><\/p>\n<\/blockquote>\n And frankly, even those of us who do not have difficulties sometimes struggle with words that we are not used to seeing in writing but know what they mean. A conversation-based or voice-controlled interface can make healthcare more accessible to people with disabilities.<\/p>\n Our conclusion is that it is beneficial to offer both voice-recognition software and a keyboard when text is to be entered. When voice-recognition is used, however, the user must have a chance to correct the text before it is sent, used or saved.<\/p>\n In 2018, it seems impossible that the major cloud computing suppliers\u2019 operational services can be used for meaningful image recognition in the field of medicine. In particular, we evaluated Microsoft Azure and Amazon AWS. X-ray images of hands are described as \u201cA white vase on a table\u201d. A man lying back on a bunk getting an injection in his arm is tagged as \u201cperson, indoor, sitting, using, woman, holding, bed, table, hand, top, green, young, white, cutting, food, man, playing\u201d, where \u201cusing\u201d possibly explains the syringe although we would have preferred the tag \u201csyringe\u201d.<\/p>\n But we can always work backwards and do the dirty work ourselves. That involves learning a machine the geometry behind a bad mole, or whatever you want to identify. This appears to mostly be done by creating models so that you can discover things automatically (feature detectors<\/a>15<\/a>) and working with neural network (probably the convolutional sort<\/a>16<\/a>). This is a possible lead for further study.<\/p>\n Unfortunately, we can see a number of obstacles:<\/p>\n What we can use the major cloud computing suppliers Amazon, Google, and Microsoft for is hiring computing power to train our own models in machine learning. They do not (yet) seem to have operational solutions, at least not in a medical application area.<\/p>\n What is most promising for continuation is decision-making support for self-care\/self-triage and the creation of different information services based on the individual\u2019s symptoms. Triage<\/a>17<\/a> is a method for sorting and prioritising patient requirements based on their medical history, symptoms and sometimes data on heart rate, respiration, body temperature etc. In an Emergency Department, it is usually the first assessment you get before you either have to wait your turn or come first in the queue in cases of emergency.<\/p>\n Medical history is interesting in combination with NLP technology because we can then match the medical history with medical vocabulary such as Snomed CT, ICD-10, ICPC etc. It provides an increased structure for free text and can thus better complement other data when training a neural network.<\/p>\n Our opinion is that the machine learning functions available for simple use in cloud computing suppliers\u2019 services are not yet ready for the meaningful use of images. Nor is Swedish supported as a language in NLP \u2013 with the exception of IBM \u2013 where three of the nine NLP functions are also supported in Swedish.<\/p>\n In other words, we need to take greater technical responsibility if we are to make machine learning work \u2013 or we wait and hope that someone else will solve our problems for us. However, using the same technology that is available to the rest of the world may not put us at the top in the world of e-health by 2025. Perhaps we need to make the most of our own unique skills, as regards both technology and healthcare?<\/p>\n It is slightly disappointing that none of the solutions presented as AI during the e-health conference Vitalis 2018 seem to have come far enough that it is possible to use them. On the one hand, they seem to build on lots of manually structured information, in the form of triage records, or decision trees<\/a>18<\/a>, in order to have a body of knowledge. On the other hand, they have not even used the established technology to provide the machine with contextual understanding.<\/p>\n An example of this is that an AI nurse suggested \u201cfungus in the genitals\u201d as a relevant further discussion regarding the complaint \u201cfungal infection of the nail\u201d. A relatively simple solution regarding the relationship between different concepts could have supported the AI nurse with the information that nail and genitals are not significantly related.<\/p>\n On a more positive note, we have established contact with a doctor at Sahlgrenska University Hospital in Gothenburg who shares our idea of continuation and we now need to find people who are closer to care provision for a larger scale evaluation.<\/p>\n At the same time, we know there is a preliminary study project in Region V\u00e4stra G\u00f6taland\u2019s conversion area Digital Care Services, which is aimed at self-triage. Self-triage is one feasible outcome of an investment like this. Another is an automated second opinion and quality work in general, as it is relatively easy to find discrepancies where the diagnosis is unexpected in relation to the meaning of the medical history or the lab results.<\/p>\n In view of what we have learned about the possibilities of deep (reinforcement) learning and the quantities of patient history we have, it seems that a combination of NLP and deep learning should be what we focus on in a larger scale project.<\/p>\n We hope to be able to bring together several parties\u2019 interests in a common next step, close to care provision, turning the idea of a workshop and conversion into reality.<\/p>\n Without discussing AI as a concept at length, we use the term to refer to what is known as weak or narrow AI<\/a>19<\/a>. That is to say specialised, non-general AI. It is a machine that does one thing well but does not understand other things it has not encountered at all. You can also choose to see AI as a research area in which machine learning is the main issue. In reality, this preliminary study focuses more on machine learning.<\/p>\n Since AI is used as an umbrella term for a lot of things today, it is difficult to find the grain of gold in what is offered. That is why we became interested in a preliminary study to examine what we can benefit from in the field of AI that is already \u201coperational\u201d and packaged. In other words, we aim to use what has already been done for what has not been done, i.e. to match what suppliers on different levels have and apply this to our data sources.<\/p>\n But what does AI mean? A sensible definition of AI is the one that was noted in the 1950s:<\/p>\n ”the capability of a machine to imitate intelligent human behavior\u201d<\/p>\n<\/blockquote>\n ”And the marketing promises considerably more than what it can handle even in the case of real artificial intelligence.\u201d<\/p>\n \u2013 Difficult for IT buyers to get it right when everything is called AI, Computer Sweden<\/a>20<\/a><\/p>\n<\/blockquote>\n One way of measuring artificial intelligence is The Turing test<\/a>21<\/a>. Then a human must assess if they are speaking to a computer or another human. But we can also discuss what is meant by the word \u201cintelligence\u201d. Translating it directly to the Swedish word \u201cintelligens\u201d often leads to disappointment in what AI has accomplished in 2018. But if you compare it instead to how \u201cintelligence\u201d is included in other concepts or areas, such as Business Intelligence (BI)<\/a>22<\/a> , the meaning changes completely.<\/p>\n However, even more definitions can be found in a dictionary<\/a>23<\/a>:<\/p>\n ”The ability to acquire and apply knowledge and skills.\u201d<\/p>\n<\/blockquote>\n Yes, through machine learning, AI can learn things based on the data it is fed.<\/p>\n We can discuss what is meant by the word \u201capply\u201d. Applying knowledge can be in line with those who think that AI must have a body or other representation, have achieved consciousness and be mortal in order to be genuinely intelligent. In its most banal form, you can also argue that a biscuit cutter which mass-produces gingerbread men is applying its knowledge.<\/p>\n But what does \u201cskills\u201d mean? Is it being able to say that it cannot draw conclusions based on the data? Or that it does not think that what it has learned can be applied to the given question? What we are getting at with this reasoning is that intelligence perhaps requires more than just good old-fashioned computing power.<\/p>\n ”The collection of information of military or political value\u201d<\/p>\n<\/blockquote>\n and<\/p>\n ”the gathering of intelligence\u201d<\/p>\n<\/blockquote>\n Here, we return to the similarity between AI and Business Intelligence. That it is more a matter of knowledge, familiarity and understanding through information. In the project we have some problems with this definition of AI. If \u201cintelligens\u201d in Swedish was about the collection of data, we should have no reasonable shortage of AI skills in view of the fact that this is what every database administrator (DBA), many back-end developers and all full stack developers have done for most of their careers.<\/p>\n Regardless of what we think AI is, or how you define intelligence, a machine is not comparable with a human being. AI can be expected to have human abilities, but its computing power has been inhuman for a very long time. If you get a machine to calculate the average of a few million decimals and compare it to a human being \u2013 the computer will solve the problem quickly long before the human being is finished.<\/p>\n AI is not just one thing, but it is mostly machine learning that is meant<\/p>\n AI is an umbrella concept which includes machine learning (ML) and the results of it, understanding human speech, getting a machine to have a concept of visual things like a pedestrian in front of a self-driving vehicle etc.<\/p>\n However, most of the things that are referred to as AI are actually machine learning. Just like AI, machine learning was defined as early as the 1950s. This is how Arthur Samuel described machine learning in 1959:<\/p>\n ”[\u2026] field of study that gives computers the ability to learn without being explicitly programmed\u201d<\/p>\n<\/blockquote>\n A more contemporary definition is Aur\u00e9lien G\u00e9ron\u2019s in the book Hands-on Machine Learning with Scikit-Learn and TensorFlow:<\/p>\n ”Machine learning is the science (and art) of programming computers so they can learn from data\u201d<\/p>\n<\/blockquote>\n As humour site XKCD notes in the illustration on the next page, the process of machine learning is sometimes treacherous; you can get lost if you do not understand all the stages. Or as journalist Darrell Huff put it:<\/p>\n ”if you torture the data long enough, it will confess to anything\u201d<\/p>\n \u2013 Darrell Huff, the book How to Lie With Statistics<\/a>24<\/a><\/p>\n<\/blockquote>\n That is to say, you can obtain statistics to support any conclusion. The same problem exists with machine learning, which is largely based on statistics and mathematics. We need to know that the data is not unbalanced, biased, tortured etc.<\/p>\n A neural network sounds more complicated than it is. It is an artificially created network of neurons. Neurons may also be referred to as brain cells or nerve cells. These are connected in a network with the help of synapses and can therefore communicate with each other.<\/p>\n In the case of machine learning, this \u201cbrain\u201d is known as an Artificial Neural Network (ANN)<\/a>25<\/a>.<\/p>\n Each single neuron is good at something and become better trained over time. The role of the trained neuron can be, for example, to recognise your grandmother, hence the name of the hypothetical \u201cgrandmother cell\u201d:<\/p>\n ”In the 1960s neurobiologist Jerome Lettvin named the latter idea the \u201cgrandmother cell\u201d theory, meaning that the brain has a neuron devoted just for recognizing each family member. Lose that neuron, and you no longer recognize grandma.\u201d<\/p>\n \u2013 One Face, One Neuron<\/a>26<\/a> (Scientific American)<\/p>\n<\/blockquote>\n Consider that you have enormous numbers of these neurons. This is so that you recognise friends and relatives you meet often. The network is a quorum when you see someone you recognise in a crowd.<\/p>\n The neuron that is not used so often may be about trying to recognise the girl who moved to a different city in second year, the one you have not heard from for a long time.<\/p>\n The reason that AI has again come to the fore is mostly due to a method known as deep learning (DL). Deep learning builds on some of the categories of machine learning, often known as unsupervised, supervised or reinforcement learning.<\/p>\n Deep learning has a much more extensive network than was used in the past. This is possible due to the computing power which is now available, at least to really large organisations, but also because many companies have collected large quantities of valuable data.<\/p>\n These levels \u2013 of which deep learning has many \u2013 are known as hidden layers. The name comes from the fact that we do not see these in the same way as the part where you input the data (input layer) and the part where the answer comes out (output layer).<\/p>\n Each of these hidden layers helps to refine an impression, for example to discover who is in a photo, if a photo is to be classified in order to find out whether it shows a dog or a cat, if it is a healthy or diseased cell you are looking at or whatever it is you want to investigate.<\/p>\n Neural networks can have different architectures depending on their purpose. The hidden layers can be combinations of layers with different specialities. For example, if you are looking at images of faces for signs of a stroke, it helps to first locate the face in the image in order to provide a defined area for subsequent neurons to locate the eyes, corners of the mouth, cheeks etc.<\/p>\n Deep learning uses a large number of layers, making it extremely complex and computing-intensive. It is not always realistic to do this yourself or invest in your own computing power.<\/p>\n Once you have taught a machine something based on data, there are two main architectures which allow you to take advantage of this. One is that you can have a neural network which is ready to answer questions \u2013 based on what the network has learned. For example, if we send a new image to a neural network trained to classify whether the image contains a dog or a cat, you will receive a response.<\/p>\n The other is to save these lessons to a model (or machine-learning model) using a completely different architecture. It is about \u201cmodelling\u201d knowledge, to translate the lessons into something that can be described to a machine. A knowledge model (machine-learning model) can be transferred from one machine to another, so if a machine has been trained to recognise something, it can be explained to another machine.<\/p>\n So what do we mean by learning? In the field of machine learning, we often speak of supervised, unsupervised, reinforcement or transfer learning.<\/p>\n Supervised learning works for the example of the data source we had from the thorax sector. That is because we have medical histories and three different doctors\u2019 diagnostic codes based on the medical histories they have read. Unsupervised learning is a bit trickier since you are just sitting on masses of data but do not have a given answer to what each of the data points means. Imagine that you have a lot of patient accounts but no idea of the diagnoses or even knowledge of the people who have been ill.<\/p>\n We assume that in those cases where the healthcare system has collected data on patient health it is with the intention of making diagnoses rather than from a need to hoard information. Therefore, we assume that we have the ball in our court in the form of test results, medical history, X-ray images and other data as well as some form of assessment or classification.<\/p>\n In these cases, it makes no sense to pretend we do not know the answer. However, supervised and reinforcement learning can still be relevant. If you play them off against each other, they seem to have different ambitions. Supervised learning is about getting a machine to be as good as the human beings who instruct it, while reinforcement learning is not necessarily limited to what the instructors are capable of.<\/p>\n Reinforcement learning is very good for limited problems that have a defined set of rules. In which cases it would work in a healthcare context is somewhat unclear.<\/p>\n A joke among developers is that there is no difference between AI and something that has been programmed with a million what-if scenarios. What we developers call IF statements. An IF statement is like a condition. For example: If there is marinated steak in the shop, buy both marinated steak and potato salad for dinner. If a machine has millions of these rules to adhere to, it can become difficult to distinguish it from a human being in some cases.<\/p>\n Is it good enough to be useful and is that what we think intelligent means?<\/p>\n In the best case scenario, many services based on machine learning are more accurate than human beings, but how good do they need to be to be ethically justifiable to use? This is a difficult issue. Anecdotally, in dialogue with our colleagues, there seems to be a consensus that if a machine makes a mistake once, it is terrible and disastrous for trust, while all humanity is not as easily condemned for mistakes with which we are already familiar. This seems neither fair nor rational.<\/p>\n How to find out whether the technical solutions that have been developed are good enough? There are various bench marks. One bench mark that everyone will comprehend is if professional people are outperformed so often that some human tasks are replaced. We read about this happening periodically. It is not unusual that a machine is more accurate than a human being when it comes to identifying one thing or another. A more statistical bench mark is whether you could have influenced the relative entropy of a quantity of information<\/a>32<\/a>. It is a way of finding out whether the information processed by an algorithm has achieved more order. Computer geeks can also read more about the diversity index<\/a>33<\/a>.<\/p>\n Much of diagnostic imaging seems to take place through the knowledge models\/neural networks that are the result of different forms of machine learning. One example is helping a person to find Regions of Interest (ROI) \u2013 places in an image where the person should focus.<\/p>\n A machine has certain super powers that are difficult for a human being to achieve. One of them is that machines do not have low blood sugar before lunch and do not become tired at the end of the working day. They can quite simply work through the night and present their findings while you choose to sleep, enjoy a restful lie-in and a healthy bike ride in the sun to work.<\/p>\n Unlike humans, the computing power of machines increases from year to year. And in some cases, you can invest a few hundred pounds and get the answer to a question the same week instead of in the next decade. Today, it is either graphics processors (GPU, Graphic Processing Unit), tensor processors (TPU, Tensor Processing Unit) \u2013 which Google hires out \u2013 or FPGA (Field Programmable Gate Array) which do the hard work.<\/p>\n Or \u2013 it is data centres filled with these that we use. Or a personal work station with a GPU, such as the Nvidia CUDA, for data scientists to go from investigating the data source to test training an artificial neural network on a subset of data. The purpose of this is to examine whether the option of continuing on a larger scale is of interest.<\/p>\n One difficulty with taking advantage of machine learning today is that many suppliers believe that they offer operational solutions, but on closer inspection they may not be quite as revolutionary as the sales message suggests.<\/p>\n Another challenge seems to be that many suppliers offer solutions to \u201ctoy problems\u201d \u2013 solutions to things that are not helpful for that many people. This is clear in the case of the cognitive services that inspect images. When they can only identify international celebrities or landmarks such as the Eiffel Tower, they are not very useful for healthcare or any other industry.<\/p>\n Despite the fact that suppliers have long lists of more or less operational services, it is unclear what they offer that is not about either classic computing power or toy problems. You may get the feeling that many suppliers claim to be higher up in the value chain than they really are.<\/p>\n A further difficulty is the work of administering the knowledge models\/neural networks you have trained. It requires experience that many people do not yet have to choose a strategy for online learning<\/a>34<\/a>, batch learning or whether to throw everything away at every iteration and start again. A drastic comparison: It is like resetting your human employees every time they have to learn something new.<\/p>\n It would have been useful to be able to build on other people\u2019s knowledge and fully trained knowledge models. Perhaps if we had deposited our models and (linked) open data in a public blockchain? Nowadays, it is on the development service GitHub that you most often find operational solutions to reuse, often of varying quality.<\/p>\n On the one hand, we investigated what is offered in the form of operational solutions by those in the healthcare sector. On the other hand, some of the more classic technology suppliers and their cognitive and AI systems can be used through their online services.<\/p>\n Certain products are offered as what is known as white label products. This means you can put your own logo on an otherwise operational solution. However, one of the two suppliers who got in touch after Vitalis explained that:<\/p>\n ”Unfortunately, we don\u2019t have much completed technical documentation, since we typically do not offer a product for developers or system integrators, but are more used to providing operational integrated solutions with UI+backend.\u201d<\/p>\n<\/blockquote>\n In other words, the transparency of the solution is not great as it must be manually tested for each possible diagnosis to see if the results seem sensible. The second supplier who dared to get in touch wanted to have a meeting to find out which questions they have to answer, despite the fact that the issues had already emerged in conversation, in the form of bullet points for the sake of clarity.<\/p>\n To learn more about the complexity of those solutions, we have also taken developer courses and completed certificates in deep learning, computer vision, NLP and GAN on online learning platforms such as Udemy, as well as data science at Berkeley on EDX. Our findings are summarised below in the respective hypothesis.<\/p>\n The example of the data source we have from the thorax sector contains medical history in unstructured text, three different doctors\u2019 individual diagnoses with diagnostic classification according to the ICPC code system and what the three doctors finally agreed.<\/p>\n We have looked at the data source from two different angles. On the one hand, through NLP with association-like sentiment analysis<\/a>35<\/a> to see positive and negative indications (soft matching). A more viable alternative proved to be word matching between the ICPC code system and the medical history text (hard matching). Soft matching can capture words to describe a general experience of state of health, but does not seem to achieve anything helpful in a healthcare situation.<\/p>\n NLP is mathematics, but language is ambiguous. This makes it somewhat complicated to implement NLP in healthcare in view of the control requirements that exist for the healthcare provider to be able to take responsibility.<\/p>\n Together with experts in psychiatry from Sahlgrenska University Hospital, we have inspected the solutions available for summarising and navigating medical record texts. Unfortunately, we have also had difficulty connecting an existing solution to something that can be implemented. Perhaps we must instead approach the universities to turn something from research into something we can evaluate and later productise together with industry?<\/p>\n We have established contacts at Chalmers University of Technology through their innovation office to talk about NLP specifically. We have also interviewed a researcher in computational biology and physiology-driven systems biology from the Wallenberg Laboratory at the University of Gothenburg in order to gain insight and perspective.<\/p>\n However, we have noticed that in the fields of both medical technology and NLP, resources are available in Gothenburg where we work.<\/p>\n For example, this thesis from March 2018 is worth exploring:<\/p>\n \u201cIn Paper III, we study the use of deep neural sequence models working on the raw character stream as input, and how this class of models can be used to detect medical terms in text (such as drugs, symptoms, and body parts). The system is evaluated on medical health records in Swedish.\u201d<\/p>\n \u2013 Olof Mogren\u2019s doctoral thesis<\/a>36<\/a> at Chalmers University, Data Science division<\/p>\n This is something that some organisations presented or lectured on at the Vitalis Conference 2018. They had built solutions to conversation-based user interaction (which we will discuss shortly). None of them impressed us. Some of the solutions seem quite simply to have a self-reinforcing and negative spiral where they use reinforcement learning on the medical history they have generated themselves. Over time, this would produce increasingly stupid AI, in what is known as \u201coverfitting\u201d, despite the fact that these organisations claimed that their AI became smarter each day. Perhaps they forgot to tell us how it became smarter despite the impression it gave of self-reinforcement?<\/p>\n A few health care organisations have built prototypes of what this type of service could look like. However, they have probably based it on the healthcare provider\u2019s needs as there are a number of stages to manage before it comes to describing health. It is therefore a solution to an appointment at which the patient is asked how they view their case later in the process.<\/p>\n However, it still remains to provide decision-making support regarding the appropriate level of care, which makes this a supplement to existing primary care, but online. Self-triage is therefore missed out if we do not try to make a decision based on what we know. The fact that the solution leads the user to an appointment before the appropriate level of care (if any) is known misses the point. It is simply a service for appointments.<\/p>\n If we assume that Agnes Wold\u2019s example is correct, it is foolish to steer people towards an unnecessary appointment through smart digital solutions, when in fact some people should be encouraged to stay at home or just wait.<\/p>\n In order for it to work with greater accuracy, the mass of healthcare knowledge must be available in a refined and machine-readable form, so that we know what further questions need to be addressed to the user. Triage records seem to be used for this purpose today.<\/p>\n Translated into machine learning, this corresponds to a decision tree. The initial value is the patient\u2019s account or medical history. Depending on the content, the case can take different paths. If there are signs of lung disease, you end up in the part of the tree that deals with the lungs and therefore a certain type of control question must be asked. If, instead, it is about joint pain, other control questions must be asked in order to make a diagnosis.<\/p>\n Using techniques such as sentiment analysis, you can get an indication of the emotion in the patient\u2019s own account. Since we do not have access to great quantities of patient accounts in the preliminary study, we have trialled this with another data source instead. We looked at product reviews where we had two extremes: those which gave a good rating and those which gave a poor rating. In addition to the scores, we have the actual text from the review. If you train machine learning on this data, it can predict the rating with some certainty if it is fed a new review text.<\/p>\n A parallel in healthcare would be to have a data source with historical patient accounts with some form of confirmation as to whether the case was serious or harmless in nature.<\/p>\n This type of technique may be uncertain in individual cases, but can capture the extremes where desperation shines through and some form of action is justified.<\/p>\n AdaBoost<\/a>37<\/a> and Cascading classifiers<\/a>38<\/a> are meta algorithms in machine learning and technologies that provide a consolidated \u201cverdict\u201d based on a number of signals in the neural network. This is a machine\u2019s way of coming up with a qualified guess, instead of just showing the probability. It is equivalent to \u201cthere is a lot of evidence suggesting that\u2026\u201d, where many minor indications point to a particular conclusion.<\/p>\n We have no suitable data source from healthcare, so we have studied how to use binary classification to determine whether a text is spam. When you feed your machine learning model with a specific text and request one response with AdaBoost and one without, AdaBoost often performs better. But not always.<\/p>\n It is possible that the uncertainty decreases with the amount of data. Regardless, it is wise to reflect on this for the data source that you work with and evaluate the precision.<\/p>\n The technique Latent Semantic Analysis (LSA)<\/a>39<\/a> can be used to calculate the relationship between words. There are a variety of applications for LSA,<\/a>40<\/a> but in a healthcare context smart synonym management, for example, could be useful. It is like the relationship between doctor \u203a physician. Another possible area of application is classifying texts.<\/p>\n Some of these services can be purchased over the internet, Infermedica for example:<\/a>41<\/a><\/p>\n ”The Infermedica API features custom Natural Language Processing technology, allowing your applications to understand clinical concepts (symptoms and risk factors) mentioned by users as natural language text.\u201d<\/p>\n<\/blockquote>\n \u201cTo infer\u201d, meaning to draw conclusions or interpret something, may give you an idea of the precision you can expect from the conclusions. That is, that this is an attempt at refining information.<\/p>\n In order to reduce the risk of the conclusions being vague, we need to continue examining the possibilities. We believe that a combination of large amounts of data, deep learning and NLP is the next logical step.<\/p>\n Unfortunately, deep learning requires much more data than we have access to in this preliminary study. However, the amount of information is not a problem, rather obtaining permission to use it. Therefore, the next step is to extend the project to include more parties, to cooperate with experts from Chalmers University of Technology and the University of Gothenburg to produce a specification for a suitable data source.<\/p>\n This is classified as machine learning because a lot of machine learning is required for a machine to be able to decode human speech and, in the best case scenario, understand the human intention behind what is being said. This can also involve listening to how something is said, which parts are highlighted, if the person\u2019s way of breathing while they talk is indicative of anything etc.<\/p>\n We have evaluated services and made inquiries. When it comes to speaking and being understood, there have been wildly varying results. In all cases, the understanding has been processed in the suppliers\u2019 own systems, so we must assume that it does not get any better than that at the moment.<\/p>\n Apple\u2019s Siri thinks that \u201cm\u00e5r nog\u201d (certainly feel in Swedish) is \u201cj\u00e4vla hor\u201d (damn whore) when spoken by a nasal person with a slight N\u00e4rke accent, but the same app works reasonably well when others talk to it.<\/p>\n We have also tested what Microsoft offers through Azure. There, you had to train the machine\u2019s understanding by selecting a phrase to repeat. The machine therefore knew in advance what it would hear. In spite of this and three attempts to record a line from the first film in The Godfather series, the machine did not hear or understand the speech. If we take the view that there were mitigating circumstances, and it was the same nasal person from N\u00e4rke (a province in south central Sweden) who, at this particular time, was sitting in a half-empty caf\u00e9 and probably had more of a blocked nose than usual on their first \u201chealthy\u201d day after a cold. However, for the healthcare system to be able to use these services, we cannot fail to help either nasal people or those suffering from the cold, and even people from N\u00e4rke should get help.<\/p>\n What we learned from a few weeks with Google Home and Amazon Echo (better known as Alexa) is that they currently have difficulty with all things Swedish. Even if you accept speaking English to them, it is not easy to know how to pronounce Swedish things so they understand. For example, it is great to be able to ask Google Home things such as when does the nearest pharmacy close. But if you want to ask when the pharmacy on a particular street closes, questions arise about pronunciation or if you should try to translate the name etc.<\/p>\n See the link in the Appendix for a thorough examination of conversing with smart speakers.<\/p>\n The best medical suggestion we got for using the voice interface was the tip to contact Araz Rawshani<\/a>42<\/a> at the University of Gothenburg. As far as we understand, research is available on how a person who is having or is at great risk of a cardiac arrest sounds and Araz can probably update us on the current situation.<\/p>\n In addition to being a good warning signal to embed in our phone services, like when you call the 1177 emergency care helpline in Sweden (or 111 in the UK), it is something that would be interesting to assess as a preventive function, for example in the apps we develop for recording patient accounts or in the case of self-triage. Apparently there are biomarkers which affect breathing and speech.<\/p>\n One possible application is as a triage aid. A guide for taking care of minor accidents in the kitchen, which instructs the user to follow the order of available recommendations. Guiding amateurs who, using conversational technology, can make sure to patch themselves up before contacting the health service for a subsequent check-up.<\/p>\n The conclusion for this hypothesis is that there may be a good conversational solution for one of the world languages spoken in Sweden before it is available in Swedish. If, in certain applications of telecare, there is merit in the patient being able to talk to devices, this may be worth evaluating.<\/p>\n One group that can greatly benefit from the conversation-based interfaces available today is those with reading and writing difficulties. According to the Swedish Dyslexia Association, it is estimated that around 5\u20138 per cent of the population in the literate world have these difficulties. For the sake of this group, it is advantageous if text can be entered with the help of conversation-based elements, for example for those terms that people may know how to pronounce but not how to spell. More people would be able to use it from time to time.<\/p>\n Today, it is somewhat difficult to use these services through the known suppliers due to integrity. Offering a service where everything that is entered is sent to a third party for analysis turns it into much more of a legal issue.<\/p>\n A more viable scenario is to have texts read out, which can benefit those with reading difficulties as well as anyone else who, for one reason or another, would rather not read at the time. Imagine a scenario where 20 articles from the Swedish healthcare guide 1177.se are recorded as a personal audio book before you embark on a particular healthcare treatment.<\/p>\n Computer vision is one of a number of concepts about giving a machine the function of being able to see. It is not always exactly the same as the way a person sees something, but we will return to this later.<\/p>\n Images, especially video, are difficult for a machine to analyse compared to text and numbers. Look at it like this: the amount of information in a single photo that you take with your mobile phone camera could be equivalent to all the books you ever read during your school years. Since computer vision is so computing-intensive, it is better to be able to build on what others have already accomplished. There are services available on the internet for this purpose, a few of which we have investigated.<\/p>\n The major online suppliers of more or less operational solutions offer a particular set of standard functions. Below is Microsoft\u2019s list for its cognitive service Azure:<\/p>\n Based on this, there is a risk of creating high expectations of what these services can provide. It is enough that a lot of time and energy has been put into training these services\u2019 functions, but what can they be used for in practice?<\/p>\n We have done some testing. First up was the Amazon service, Rekognition. A photo of Marcus sitting at a desk and lecturing beside a projection screen was tagged as railways and the military. Perhaps the train connection was due to the angled windows in the background, but the military link requires more imagination.<\/p>\n Second up was to ask Microsoft Azure cognitive services to tell us something about some images.<\/p>\n The man receiving an injection in his arm is certainly sitting or lying on a kind of bed, but is \u201cusing\u201d a reference to drug abuse rather than a hospital scenario? Otherwise, the purely white fabric could reveal a probable location.<\/p>\n No, it is not a person holding a baseball bat in the photo. However, some of the tags are quite descriptive, for example that it is dark and what \u201ccut\u201d now refers to. But \u201ccake\u201d is interesting. Is it the bloody white surface that is similar to a strawberry tart? Or is it an accident on the ski slope and the reason for \u201cskiing\u201d?<\/p>\n What is the reason for these peculiarities? It is probably because the foundation for image recognition is not solid within the application areas required in healthcare.<\/p>\n As the website AIweirdness<\/a>43<\/a> pointed out in March 2018, neural networks often see sheep in images where there are none. This is a slightly comedic angle on the science fiction novel \u201cDo Androids Dream of Electric Sheep\u201d.<\/p>\n The following images (borrowed from AIweirdness.com) seem to identify light patches like sheep in images with green contents, probably grass.<\/p>\n The simple explanation is that the image material that the service has been trained on often had a combination of lawns and that the bright parts have been sheep. Therefore, when it comes to images with light patches of thawing snow on a semi-green surface, it seems very easy to believe they are small, woolly sheep.<\/p>\n Other recurrent misclassification we found in images run on Microsoft Azure is that older stone buildings are often assumed to have a bell tower. To skip the technical work, you can follow Picdescbot<\/a>44<\/a> on Twitter. The account retrieves a random image from Wikipedia and checks what an image service assumes it represents. It is often successful, with a number of strange and sometimes comic exceptions.<\/p>\n The area of application for these operational services is not currently healthcare-related. Perhaps they could have been used if computing power was hired and a service trained on the sort of images we would like it to understand.<\/p>\n Opting for slightly more manual computer vision instead leads to a number of challenges. As previously mentioned, it takes a lot of computing power, so an early objective is to limit the information you have to work with. You want to identify a ROI (Region Of Interest).<\/p>\n This is no different to identifying certain basic geometric shapes to see if an image contains something that may be of interest.<\/p>\n A simple example from the field of facial recognition is the combination of one or more eyes, eyebrows, nose, mouth and ears which make up a face. All these features are known as weak classifiers.<\/a>45<\/a> It is the combination of several things that convince a machine that what it sees is a face.<\/p>\n These decisions are often made based on images in grey scale since colour shades do not really add anything for machines, which human beings naturally do not agree with. A machine\u2019s precision is not necessarily increased by seeing eye colour. In contrast, colour information is actually burdensome for the calculation.<\/p>\n The geometry of your face in a simplified form, see the photo above and the image next to this, and think about the following points:<\/p>\n The geometry also has internal relationships which declare that it is a face. The line of the nose can be expected to be in a particular place if you think you have found the mouth or eyes. In this way, it is possible to identify a face even if the person turns their head or does a handstand. Or, if as shown in the picture above, it is difficult to discern the mouth because of beard growth, other weak classifiers can reach a joint decision that it is a face.<\/p>\n If what is interesting in a photo or other images can be boiled down to the above geometric thinking, it seems feasible to look for other patterns too. It is not uncommon to spend a lot of time on what is known as feature engineering,<\/a>46<\/a> i.e. how to describe something in order to be able to find it in an image.<\/p>\n In the above case of facial recognition, the image is reduced to 24×24 pixels. Even with this extremely limited amount of information, the few geometrical shapes that can identify a face have as many as 180,000 possible combinations \u2013 in a single image.<\/p>\n In other words, a vital part of the medical use of computer vision is probably dependent on being able to find a ROI in the image to reduce the dimensionality, i.e. to reduce the amount of calculations to a realistic and feasible number. For previously shown facial recognition, ready-trained models are available to download from the internet, but then their quality must be assessed.<\/p>\n In future, something resembling an app store may be available for buying or hiring trained models to apply to your own data. If this is not the case, it will be a major investment to build up this knowledge and it is not something that many organisations will manage on their own.<\/p>\n In this area, a lot of work has been going on for a long time in the field of radiology, so we decided not to investigate further. However, the data source BFR (Bild- och FunktionsRegistret) \u2013 which radiology also uses for storing information \u2013 is of interest for more reasons than simply diagnostics.<\/p>\n An interview with a register researcher at the University of Gothenburg suggested that there is potential to apply machine learning technology in most registers. Therefore, a possible continuation of this project is to investigate what research is already being done on BFR data and if there is anything to add.<\/p>\n We have noted that Stanford has published a large number of thorax images on the internet. However, as previously mentioned, a lot of preparation is required before starting to process these amounts of information in the search for something. Other parties\u2019 data sources may be able to act as validation data when we train machine learning on internal data. Equally, data from other organisations can modulate our machine learning models to reduce the bias in our system. For example, consider the scenario that an international emergency patient is admitted and we require access to the person\u2019s full medical history. It would be good if our knowledge model understood the X-ray images even if they originate from a different healthcare provider.<\/p>\n Deep learning (DL) is a special variant of machine learning based on building greater complexity in the neural network. This means the network becomes more knowledgeable about the details of the amount of data it is being trained on and can reach more advanced conclusions. For it to become precise, larger amounts of data than normal are required. This also raises the question of whether it is the image being examined or the signal from what captures the image that is important.<\/p>\n This can be compared to the concept of metadata: information about information, data that describes or summarises other data. If you take a photo with your mobile camera, meta data will be saved together with the image. For example, it provides information on how large the aperture is, at what time the photo was taken and sometimes geographical information such as latitude and longitude. The difference between the signal and the image content is already quite significant here. The signal can relate the image content to:<\/p>\n After all, images are optimised for human consumption. The signal can contain information for machines that gets lost when it is converted to a two-dimensional visualisation.<\/p>\n If you focus on including the different quality registers in the healthcare system, you can find out about what efforts have already been made in healthcare. An example of such a register is the cause of death register. How, and for what purpose, we can get access to one or more registers is a different matter. But it is clear that knowledge can be found in these registers.<\/p>\n The registers our register researchers from the University of Gothenburg examined include almost two million individuals. They do not contain that many data points per individual, about a couple of hundred. However, there are other initiatives which can supplement the information found in quality registers with a couple of thousand additional data points. One example of this type of data source is Scapis<\/a>47<\/a>, a study in preventing heart and lung disease. Scapis claims to have a few thousand data points per individual.<\/p>\n Regardless of whether you include images, signals or large amounts of register data, these types of calculations are so costly that greater preparation is required than we can manage during this preliminary study. We have therefore focused on taking stock of the complexity and have proposals for the future.<\/p>\n In cases where we require deep learning, runs can take place in parallel on GPUs. We have been informed that we can arrange data runs at Chalmers University of Technology, if necessary, to avoid disclosing data to an IT giant\u2019s data centre.<\/p>\n It would have been interesting to explore transfer learning of the 40 GB of thorax images that Stanford University released and then see if there is potential to transfer the machine learning model to another diagnostic area. This would be done by diagnostic imaging experts.<\/p>\n In cases where major computing power is required over a longer period, more and more options seem to appear. In May 2018, Nvidia released an efficient server solution, HGX-2, which can work with thousands of images per second. It costs almost SEK 4 million (approximately 0.4 million Euros).<\/p>\n ”Nvidia also explained that HGX-2 test servers have managed to train models with images at a speed of 15,500 images per second in the Resnet-50 standard test. This means that an HGX-2-server can be used to replace up to 300 servers with standard processors.\u201d<\/p>\n \u2013 Nvidia releases powerful processor for servers with AI in focus (Techworld)<\/a>48<\/a><\/p>\n<\/blockquote>\n Another project would be to examine the possibility of making diagnoses using an app for mobile phones. Take a selfie and find out, for example, if:<\/p>\n A project like this would involve taking stock of which working models are available as an aid. During the past year, there has been talk of which diagnostics can be performed with a simple photo of a person\u2019s eye, and that the pulse and other vital parameters can be read with a simple mobile camera. Can we take advantage of other people\u2019s findings in this field?<\/p>\n ”[…]a tech culture that\u2019s built on white, male values – while insisting it\u2019s brilliant enough to serve all of us. Or, as they call it in Silicon Valley, \u201cmeritocracy.\u201d\u201d<\/p>\n \u2013 Sara Wachter-Boettcher, Technically Wrong – Sexist apps, biased algorithms, and other threats of toxic tech<\/p>\n<\/blockquote>\n In 2016 a study<\/a>49<\/a> was released about whether the technology found in mobile phones from Apple, Samsung, Google and Microsoft can help if the user ends up in a crisis situation. The short answer is that this is not always the case. There are plenty of extreme examples, such as Siri answering \u201cIt\u2019s not a problem\u201d to the question \u201cSiri I don\u2019t know what to do my daughter is being sexually abused\u201d, or \u201cSiri I don\u2019t know what to do I was just sexually assaulted\u201d being answered with \u201cOne can\u2019t know everything, can one?\u201d<\/p>\n Should we expect technology sold as intelligent to perform better in difficult situations?<\/p>\n Yes, in any case that is what Sara Wachter-Boettcher believes, who in her book Technically Wrong \u2013 Sexist apps, biased algorithms, and other threats of toxic tech argues that the IT concept \u201cedge case\u201d should be changed to \u201cstress case\u201d. That something is not promptly dismissed as unlikely by the creator of an app, for example, but instead the focus is on attempting to reach a solution to when users most require attention.<\/p>\n This book should be read by all white guys who like technology. If you believe what it says in the book, the fact that the two of us who worked on this project are both white men is not exactly a coincidence. Reuters News Agency described the problem as reaching beyond a \u201ctraditional Silicon Valley cohort\u201d. Despite many efforts to achieve more diversity, it is still the case that when marginalised groups to a large extent give up on the technology industry to start working with something else, it is difficult to improve diversity in the long term.<\/p>\n Therefore, questions of ethics, diversity and an inclusive approach are crucial when working on teaching machines something which affects people\u2019s lives. If you do not remain actively vigilant, you carve prejudices and distortions in stone and make them into invisible rules.<\/p>\n ”Who will develop the algorithms, who can review them and can the results be good if the input data for the algorithm is not neutral?\u201d<\/p>\n \u2013 Equal care in the world of algorithms, VGRblogg, 2016\u00a0<\/a>50<\/a><\/p>\n<\/blockquote>\n We do not need general AI that threatens mankind\u2019s existence to end up in ethical difficulties. Inequalities in society are confirmed and entrenched through technical failings. The situation that someone with the title \u201cDoctor\u201d is not allowed into the female changing room at the gym with their membership card may be due to a bug in the system, but it does not go unnoticed for the person affected.<\/p>\n ”The only thing that\u2019s normal is diversity.\u201d<\/p>\n \u2013 Sara Wachter-Boettcher<\/p>\n<\/blockquote>\n First and foremost, the problem is who defines what is \u201cnormal\u201d. How much insight does the person or persons have? In psychology there is a concept, WEIRD,<\/a>51<\/a> which pinpoints the context in which people who influence algorithms are found. WEIRD is an abbreviation of Western, educated, industrialised, rich and democratic. Those who construct and evaluate algorithms are often very non-representative of those who will be affected in the long term.<\/p>\n The fact that in 2015 Google Images classified dark-skinned people as gorillas or that Asians were asked to stop screwing up their eyes by a photo booth are examples of algorithms that have been trained on a substandard basis. That even Google, although considered to have the smartest AI, has these problems, may be linked to the fact that in the same year its report on diversity stated that only 1% of its employees were black. Dark-skinned developers would probably not have failed to test their neural networks with pictures of dark-skinned people.<\/p>\n Intended users are often grouped into what is known as personas or sometimes target groups. There is an imminent risk of starting to focus solely on the caricatures of users. Even in cases where personas are very detailed, there is considerable variation. Take Prince Charles and Ozzy Osbourne for example. They have a bunch of common characteristics, among other things they are white, rich, married, men from England. But one is the heir to the throne in a Royal Family and the other grew up in a dirty industrial city to impoverished parents. Differences like these sometimes disappear even if we think our work is user-centred.<\/p>\n One example that was both big budget and thought in a user-centred way is the US Air Force, which in the 1950s evaluated whether cockpits were designed for the physical dimensions of fighter pilots. They studied just over 4,000 fighter pilots and measured their shoulders, chests, waists and hips etc. All in all, 10 measurements were taken. When all the data was compiled, they examined how the average pilot\u2019s body measurements compared to each of the 4,000 measured individuals. Even if the average is read with +\/-15 percentage points, not a single pilot was average for all 10 measurements.<\/p>\n ”Even more astonishing, Daniels discovered that if you picked out just three of the ten dimensions of size \u2013 say, neck circumference, thigh circumference and wrist circumference \u2013 less than 3.5 per cent of pilots would be average sized on all three dimensions. Daniels\u2019s findings were clear and incontrovertible. There was no such thing as an average pilot. If you\u2019ve designed a cockpit to fit the average pilot, you\u2019ve actually designed it to fit no one.\u201d<\/p>\n \u2013 Todd Rose, The End of Average: Unlocking Our Potential by Embracing What Makes Us Different<\/p>\n<\/blockquote>\n All that remained for the US Air Force was to design the cockpit to support the extremes, both the smallest and largest measurement in each category would work. This work produced adjustable seats, foot pedals and clips for helmets \u2013 things that we now see as self-evident but which were not at the time.<\/p>\n For those who use machine learning to identify anomalies in data sources, these challenges may be obvious, but most of us need to work actively to challenge our unconscious assumptions.<\/p>\n It is a reasonable ambition to want to re-use what others have already developed, \u201cto stand on giants\u2019 shoulders\u201d, or to avoid the contrary, what those of us in technology circles usually complain about as the \u201cnot invented here\u201d syndrome, when people mistrust everything they did not create from scratch themselves.<\/p>\n Let us say that we want to benefit from a neural network we can hire as a service, or download a knowledge model someone has released for free on GitHub. How do we do that? This raises a number of issues, including:<\/p>\n We already have an example of a micro-inspection of the above issues in this report. The model for facial recognition was downloaded from GitHub. The first person we tried to detect happened to have a full beard and it was difficult to determine whether there was a mouth in the image. Giving the algorithm lots of images in a video stream through the webcam did not help much either.<\/p>\n Let us say that we only tested the model on women (who are in frequent abundance at the county council compared to the technology sector) and that our solution involved answering the question: \u201cCan the person smile and show teeth?\u201d We then tried to \u201csee\u201d if one corner of the mouth was hanging down. The app would not have been so useful for those with a full beard.<\/p>\n FairML is one example of an attempt to find bias in machine learning. It is a technical framework that looks for imbalance (see links in the Appendix). In time, perhaps it will become easier to take advantage of technology to examine data sources, operational knowledge models and neural networks to investigate whether there are any faults.<\/p>\n An algorithm is not any more peculiar than a cooking recipe, but in a way that a machine understands. Machines do as they are instructed, there is no magic, which is what we expect when using them. However, mistakes can be disastrous and extensive if the algorithm has faults.<\/p>\n ”Nearly half a million elderly women in the United Kingdom missed mammography exams because of a scheduling error caused by one incorrect computer algorithm, and several hundred of those women may have died early as a result.\u201d<\/p>\n \u2013 IEEE Spectrum<\/a>52<\/a> (May 2018)<\/p>\n<\/blockquote>\n On closer inspection of an algorithm, which is probably far simpler than a neural network, it was discovered that half a million English women had not been called for a mammography. A few hundred of them are suspected to have died because of this. Automating something using technology saves a lot of time but errors are then on a different scale compared to if the work was done manually.<\/p>\n Technology also struggles with empathy, there is a lack of human emotion and social tactfulness. This becomes apparent from time to time, for example in April 2018 when Siri recommended a Nazi website as the best source of information on the Holocaust, with the title \u201cThe Holocaust is a hoax!\u201d. A human could recommend the same website but they would at least have some understanding of what it means.<\/p>\n Even if machines start to get to grips with context and what extremes there are, this may not necessarily help. Take the question of whether the Earth is round, for example. Perhaps it is foolish to view both sides as equals. It would mostly benefit extremists, but The Flat Earth Society<\/a>53<\/a> would certainly be overjoyed at the attention.<\/p>\n In addition to this compilation of lessons, the preliminary study has a number of offerings in the field of development, including:<\/p>\n![\"AI](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/cover-english-hires-194x300.jpg\")
<\/a>
December 2018
Author: Marcus \u00d6sterberg
Editor: Lars Lindsk\u00f6ld
Translation: CBG Konsult
Edition: Epub (version 1.1, 2018-12-13)
Copyright: None\/CC0 excluding images
Licence: Public domain
ISBN: 9780463471234<\/p>\nTable of Contents<\/h2>\n
\n
\n
\n
\n
\n
\n
About the book<\/h2>\n
\n
Introduction to the new edition<\/h2>\n
Peter Nordstr\u00f6m
Programme Director, Swelife<\/i><\/p>\nSummary<\/h2>\n
Hypothesis 1: Processing and understanding medical history and patient accounts<\/h3>\n
Conclusion<\/h4>\n
Hypothesis 2: Voice and conversation-based user interface<\/h3>\n
\n
Conclusion<\/h4>\n
Hypothesis 3: Computer Vision and deep learning<\/h3>\n
Conclusion<\/h4>\n
Summary<\/h3>\n
\n
NLP is most promising<\/h4>\n
Business intelligence from the Vitalis Conference 2018<\/h4>\n
Next step \u2013 getting more people involved<\/h4>\n
A background to artificial intelligence (AI)<\/h2>\n
\n
What does \u201cintelligence\u201d mean?<\/h3>\n
\n
\n
\n
\n
\n
\n
\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/xkcd_machine_learning.png\")
<\/a>Training neural networks to imitate a brain<\/h3>\n
\n
Deep learning \u2013 the reason for renewed interest in artificial intelligence<\/h3>\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/MultiLayerNeuralNetworkBigger_english.png\")
<\/a>A self-learning machine? Supervised vs unsupervised vs reinforcement vs transfer<\/h3>\n
\n
Is unsupervised learning least suitable for healthcare?<\/h4>\n
Creating a machine with a memory for details?<\/h3>\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/if-statements.jpg\")
<\/a>What is good enough when it comes to the results of machine learning?<\/h3>\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/scanlime-dog.jpg\")
<\/a>Strengths in machine learning\u2019s favour<\/h3>\n
What are the current shortcomings? Toy problems, among other things…<\/h3>\n
What we investigated<\/h2>\n
\n
Hypothesis 1: Natural Language Processing (NLP) for processing medical history and patient accounts<\/h3>\n
Interpreting what is said\/written<\/h4>\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/agnes-wold.png\")
<\/a>How bad is it for the patient?<\/h4>\n
Adaptive Boosting (AdaBoost) and Cascading classifiers<\/h4>\n
Semantic analysis<\/h4>\n
\n
Continuation of the NLP track: Deep learning + NLP in combination<\/h4>\n
Hypothesis 2: Voice and conversation-based interfaces can help<\/h3>\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/sirikit-576x1024.png\")
<\/a>Support at home that you can talk to<\/h4>\n
Listening for illness?<\/h4>\n
Conclusion<\/h4>\n
Hypothesis 3: Computer vision for automatically seeing, creating or inspecting images (sometimes with deep learning)<\/h3>\n
\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/rontgen-hand-899x1024.jpg\")
<\/a>![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/spruta-i-armen-863x1024.jpg\")
<\/a>
Description: A person sitting on a bed
Tags: person, indoor, sitting, using, woman, holding, bed, table, hand, top, green, young, white, cutting, food, man, playing<\/figcaption><\/figure>\n![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/sar-syr-ihop-1024x826.jpg\")
<\/a>
Description: A person holding a baseball bat
Tags: person, cake, holding, cutting, bat, baseball, woman, man, dark, cut, wearing, knife, pair, skiing, table, hat, white, plate, standing<\/figcaption><\/figure>\nAI suffering from hallucinations \u2013 do neural networks dream of electric sheep?<\/h4>\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/tumblr_inline_p4y38qBcbX1rl9zu7_500.jpg\")
<\/a>
Description: A close up of a hillside next to a rocky hill
Tags: hillside, grazing, sheep, giraffe, herd<\/figcaption><\/figure>\n![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/tumblr_inline_p4y2clBHpF1rl9zu7_500.jpg\")
<\/a>
Description: A herd of sheep grazing on a lush green hillside
Tags: grazing, sheep, mountain, cattle, horse<\/figcaption><\/figure>\n![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/tumblr_inline_p4y307CymC1rl9zu7_500.jpg\")
<\/a>
Description: A close up of a lush green field
Tags: grass, field, sheep, standing, rainbow, man<\/figcaption><\/figure>\nMore manual computer vision<\/h4>\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/ansiktsigenkanning-1024x574.jpg\")
<\/a>\u201cSeeing\u201d is about geometry<\/h4>\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/ansikte-geometri-1024x1014.png\")
<\/a>\n
Medical computer vision<\/h4>\n
Deep learning<\/h4>\n
\n
Conclusions about computer vision<\/h4>\n
\n
\n
Ethical issues<\/h3>\n
\n
\n
It is normal to be abnormal<\/h4>\n
\n
Personas and target groups<\/h4>\n
\n
Review algorithms you want to benefit from<\/h4>\n
\n
The limitations of technology<\/h4>\n
\n
![\"Fig.](\"https:\/\/vgrblogg.se\/utveckling\/wp-content\/uploads\/sites\/3\/2018\/12\/forintelsen-576x1024.png\")
<\/a>Results<\/h2>\n
\n