{"id":4629,"date":"2024-10-08T09:55:57","date_gmt":"2024-10-08T07:55:57","guid":{"rendered":"https:\/\/www.rivistaeco.com\/?p=4629"},"modified":"2024-10-08T09:55:57","modified_gmt":"2024-10-08T07:55:57","slug":"the-marriage-between-ai-and-medicine-is-built-on-trust","status":"publish","type":"post","link":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/","title":{"rendered":"The Marriage Between AI and Medicine is Built on Trust"},"content":{"rendered":"

Artificial intelligence can significantly improve the accuracy of diagnoses and the early detection of diseases. However, many barriers remain to its full utilisation in medicine. As various studies show, trust in technology must be built by involving doctors in the development process.<\/span><\/i><\/p>\n

With the increase in life expectancy and, consequently, the elderly population, the demand for medical services is growing, clashing with the shortage of doctors and healthcare personnel. AI can offer innovative solutions and crucial contributions to medical practice. In the not-so-distant future, we can imagine a close collaboration between healthcare staff and AI tools, much like airplane pilots have been operating flights alongside algorithms for years.<\/span><\/p>\n

Great Potential but Limited Use of AI<\/b><\/h3>\n

Many aspects of medical practice require making predictions, including diagnosis and treatment. AI algorithms can contribute to data management, the development of personalised care plans, and support for diagnostic decisions and therapeutic treatments. AI can improve diagnostic accuracy, early disease detection, and reduce the time spent on routine tasks, allowing for a deeper doctor-patient relationship. It can also offer crucial support to scientific research by processing large amounts of data. AI technologies have been used for drug and vaccine discovery, including the development of COVID-19 vaccines. Recently, the development of the AI algorithm AlphaFold 3 was announced, which can predict the three-dimensional structure of proteins with unprecedented accuracy and the interactions between all molecules in living cells, paving the way for new drugs and treatments.<\/span><\/p>\n

Despite its great potential, AI adoption in medicine remains low. In a study published in the <\/span>American Economic Review Papers and Proceedings<\/span><\/i>, Avi Goldfarb, Bledi Taska, and Florenta Teodoridis considered all online job postings for hospital positions in the USA between 2015 and 2018: only about 1 in 1250 postings required AI skills, the lowest adoption rate in any industry requiring qualified personnel.<\/span><\/p>\n

The limited use of AI in medicine is due to various barriers. First, algorithmic barriers: the lack of transparency in how the algorithm works and difficulties in interpreting its results, as well as the production of biased predictions and recommendations. In a famous study published in <\/span>Science<\/span><\/i>, Ziad Obermeyer and co-authors discuss racial bias in a widely used diagnostic AI algorithm in the USA: it assigned the same risk level to African American patients even though they were in much worse health conditions compared to white patients, leading to the number of African American patients identified for additional care being reduced by more than half. Moreover, white patients had a socioeconomic background that led them to undergo more preventive and screening tests, providing more data that helped the algorithm assign the risk score. In another article published in <\/span>Science<\/span><\/i>, Samuel Finlayson and co-authors discuss examples of what they call the “dark side of AI”: changing a few pixels in an image of a benign lesion or simply rotating the image can lead AI to identify it as malignant; labelling a patient’s condition as “alcohol abuse” or “alcohol dependence” can change the diagnosis produced by AI systems.<\/span><\/p>\n

Second, legal and data collection barriers: there are no clear rules on who owns health data, who can use it, and the division of responsibilities between the algorithm developer and the doctor using it. Additionally, there are difficulties in using health data o train AI models without violating privacy rights.<\/span><\/p>\n

Finally, trust barriers in AI’s usefulness: the decision to use an AI algorithm in medical practice ultimately depends on trust in the technology. Low trust can be due to the scarcity of validation data outside the laboratory or the absence of collaboration between AI developers and clinical experts. Doctors may also feel threatened by AI, fearing it could render their work obsolete or expose them to liability for diagnostic and therapeutic errors due to predictions produced by an algorithm.<\/span><\/p>\n

Building Trust in Technology<\/b><\/h3>\n

The issue of trust hindering the use of AI in the medical field is central. As always with new technology, trust is generally low. When applied to healthcare, an instinctive attitude of suspicion and heightened attention prevails. Many studies have analysed this phenomenon, particularly among radiologists, the first specialty to exploit AI’s computational capabilities.<\/span><\/p>\n

In the most recent study, conducted by the European Society of Radiology in 2022 on a sample of radiologists with direct clinical experience with an AI algorithm, 76% considered the results produced by AI reliable, but only 37% cited the use of AI in diagnostic reports, and only 17% informed patients of having used an AI algorithm. Seventy percent stated they did not benefit from a significant reduction in workload. In an experiment that randomly assigned the availability of AI tools to professional radiologists, Nikhil Agarwal and co-authors compared diagnoses made with only an X-ray, diagnoses with X-ray and patient’s medical history, and diagnoses with X-ray, patient’s medical history, and AI algorithm prediction. The results show that cognitive biases prevent radiologists from optimally using the information provided by algorithms, and the most accurate decisions are those made either only by radiologists or only by AI. The study concludes that the use of AI is not contraindicated in radiology but that its use requires training and the acquisition of specific technical knowledge to allow the technology to achieve faster and more accurate diagnoses.<\/span><\/p>\n

A study by Peter Winter and Annamaria Carusi on the development of AI for early diagnosis of pulmonary hypertension, published in <\/span>Science and Technology Studies<\/span><\/i>, demonstrates that trust in technology is not something that arises at the end of its development and application process; rather, it is born throughout the technology’s design process, is “situated” in working practices among different subjects, and is the result of the “invisible work” behind the technology.<\/span><\/p>\n

This type of analysis counterbalances the tendency to view AI as a “technological fix” (a solution produced by technology) in the hope of solving concrete problems such as human error in interpreting thousands of mammograms or dermatological images. Instead, it highlights how social relationships (the involvement of stakeholders and clinicians) and human labour (data labelling) are part of the social construction process of the technology. In particular, it was found that collaboration practices between doctors and algorithm developers are important for validating the latter, thereby generating widespread trust in the approved AI tool.<\/span><\/p>\n

The Italian Study<\/b><\/h3>\n

In this direction, within the PNRR Fair (Future Artificial Intelligence Research) project and in collaboration with the National Institute of Nuclear Physics (INFN), we are investigating the pervasiveness of AI methods in medical practice and their impact on the professional practices of doctors and medical physicists. It is the first study to investigate trust, use, and perspectives of AI use by different categories of doctors: radiologists, neurologists, nuclear medicine doctors, oncologists, and radiotherapists, based on a sample of doctors working in Italian hospitals and research centers. Its results will be used to define guidelines for the effective development and efficient use of AI tools in medical practice through a co-creation, participation, and validation approach.<\/span><\/p>\n

Preliminary results show optimism and trust in AI: doctors expect positive effects on the economy, society, their quality of life, as well as a reduction in work times and increased diagnostic accuracy. Forty-five percent of the survey participants have used an AI system in medical practice, and 68% believe in the complementarity between humans and AI algorithms. An innovative element of the study is the use of a randomised experiment to test the potential use and effectiveness of an AI software under development by INFN for reporting, monitoring, and therapeutic guidance in patients with glioblastoma (a malignant brain tumor). The experiment’s results show that the availability of a randomised clinical validation study of the software positively affects both\u00a0 expectations of its use to reduce work times and the likelihood of using the algorithm’s results even if they conflict with one’s own diagnosis. The likelihood of accepting an algorithm’s diagnosis when it diverges from one’s own increases when a participatory approach is used in AI software development directly involving clinicians.<\/span><\/p>\n

The main strategic recommendations emerging from our study are twofold: use randomised experiments to test the validity of AI tools and adopt a participatory approach in the AI development process involving experts and end-users (doctors and healthcare personnel) directly. These recommendations align with those promoted in the “Research Services Report” of the European Parliament on AI use in medical practice. The Report emphasises the importance of a “co-creation” approach between designers and AI system users in medicine as the first step in creating an “AI passport” to ensure a traceability process that contributes to both transparency and trust in AI use in medical practice.<\/span><\/p>\n

Bio<\/b><\/em><\/p>\n

Chiara Binelli is an associate professor of economics at the University of Bologna and co-director of the Center for Research and Social Progress.<\/span><\/em><\/p>\n

Laura Sartori is an associate professor of sociology at the University of Bologna where she studies the social and political implications of artificial intelligence.<\/span><\/em><\/p>\n","protected":false},"excerpt":{"rendered":"

Artificial intelligence can significantly improve the accuracy of diagnoses and the early detection of diseases. However, many barriers remain to its full utilisation in medicine. […]<\/p>\n","protected":false},"author":7952,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[1],"tags":[],"coauthors":[159,160],"class_list":["post-4629","post","type-post","status-publish","format-standard","hentry","category-non-categorizzato"],"acf":[],"yoast_head":"\nThe Marriage Between AI and Medicine is Built on Trust - Rivista Eco<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"The Marriage Between AI and Medicine is Built on Trust - Rivista Eco\" \/>\n<meta property=\"og:description\" content=\"Artificial intelligence can significantly improve the accuracy of diagnoses and the early detection of diseases. However, many barriers remain to its full utilisation in medicine. […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/\" \/>\n<meta property=\"og:site_name\" content=\"Rivista Eco\" \/>\n<meta property=\"article:published_time\" content=\"2024-10-08T07:55:57+00:00\" \/>\n<meta name=\"author\" content=\"Chiara Binelli, Laura Sartori\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Chiara Binelli, Laura Sartori\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"8 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/2024\\\/10\\\/08\\\/the-marriage-between-ai-and-medicine-is-built-on-trust\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/2024\\\/10\\\/08\\\/the-marriage-between-ai-and-medicine-is-built-on-trust\\\/\"},\"author\":{\"name\":\"Chiara Binelli\",\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/#\\\/schema\\\/person\\\/0082e96dca462bcfc43bb541b2cd09b8\"},\"headline\":\"The Marriage Between AI and Medicine is Built on Trust\",\"datePublished\":\"2024-10-08T07:55:57+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/2024\\\/10\\\/08\\\/the-marriage-between-ai-and-medicine-is-built-on-trust\\\/\"},\"wordCount\":1482,\"articleSection\":[\"Non categorizzato\"],\"inLanguage\":\"en-US\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/2024\\\/10\\\/08\\\/the-marriage-between-ai-and-medicine-is-built-on-trust\\\/\",\"url\":\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/2024\\\/10\\\/08\\\/the-marriage-between-ai-and-medicine-is-built-on-trust\\\/\",\"name\":\"The Marriage Between AI and Medicine is Built on Trust - Rivista Eco\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/#website\"},\"datePublished\":\"2024-10-08T07:55:57+00:00\",\"author\":{\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/#\\\/schema\\\/person\\\/0082e96dca462bcfc43bb541b2cd09b8\"},\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/2024\\\/10\\\/08\\\/the-marriage-between-ai-and-medicine-is-built-on-trust\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/2024\\\/10\\\/08\\\/the-marriage-between-ai-and-medicine-is-built-on-trust\\\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/2024\\\/10\\\/08\\\/the-marriage-between-ai-and-medicine-is-built-on-trust\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"http:\\\/\\\/www.rivistaeco.com\\\/en\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"The Marriage Between AI and Medicine is Built on Trust\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/#website\",\"url\":\"https:\\\/\\\/www.rivistaeco.com\\\/\",\"name\":\"Rivista Eco\",\"description\":\"\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.rivistaeco.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/www.rivistaeco.com\\\/#\\\/schema\\\/person\\\/0082e96dca462bcfc43bb541b2cd09b8\",\"name\":\"Chiara Binelli\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/6efa21323c55820fcce3e903fc80e9908b0b069289ee8cee40b589f840d8bbf8?s=96&d=mm&r=g77a73fa6a8661705b0ff8a3d2fe5887a\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/6efa21323c55820fcce3e903fc80e9908b0b069289ee8cee40b589f840d8bbf8?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/6efa21323c55820fcce3e903fc80e9908b0b069289ee8cee40b589f840d8bbf8?s=96&d=mm&r=g\",\"caption\":\"Chiara Binelli\"},\"url\":\"https:\\\/\\\/www.rivistaeco.com\\\/en\\\/author\\\/cbinelli\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"The Marriage Between AI and Medicine is Built on Trust - Rivista Eco","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/","og_locale":"en_US","og_type":"article","og_title":"The Marriage Between AI and Medicine is Built on Trust - Rivista Eco","og_description":"Artificial intelligence can significantly improve the accuracy of diagnoses and the early detection of diseases. However, many barriers remain to its full utilisation in medicine. […]","og_url":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/","og_site_name":"Rivista Eco","article_published_time":"2024-10-08T07:55:57+00:00","author":"Chiara Binelli, Laura Sartori","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Chiara Binelli, Laura Sartori","Est. reading time":"8 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/#article","isPartOf":{"@id":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/"},"author":{"name":"Chiara Binelli","@id":"https:\/\/www.rivistaeco.com\/#\/schema\/person\/0082e96dca462bcfc43bb541b2cd09b8"},"headline":"The Marriage Between AI and Medicine is Built on Trust","datePublished":"2024-10-08T07:55:57+00:00","mainEntityOfPage":{"@id":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/"},"wordCount":1482,"articleSection":["Non categorizzato"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/","url":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/","name":"The Marriage Between AI and Medicine is Built on Trust - Rivista Eco","isPartOf":{"@id":"https:\/\/www.rivistaeco.com\/#website"},"datePublished":"2024-10-08T07:55:57+00:00","author":{"@id":"https:\/\/www.rivistaeco.com\/#\/schema\/person\/0082e96dca462bcfc43bb541b2cd09b8"},"breadcrumb":{"@id":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.rivistaeco.com\/en\/2024\/10\/08\/the-marriage-between-ai-and-medicine-is-built-on-trust\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"http:\/\/www.rivistaeco.com\/en\/"},{"@type":"ListItem","position":2,"name":"The Marriage Between AI and Medicine is Built on Trust"}]},{"@type":"WebSite","@id":"https:\/\/www.rivistaeco.com\/#website","url":"https:\/\/www.rivistaeco.com\/","name":"Rivista Eco","description":"","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.rivistaeco.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Person","@id":"https:\/\/www.rivistaeco.com\/#\/schema\/person\/0082e96dca462bcfc43bb541b2cd09b8","name":"Chiara Binelli","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/6efa21323c55820fcce3e903fc80e9908b0b069289ee8cee40b589f840d8bbf8?s=96&d=mm&r=g77a73fa6a8661705b0ff8a3d2fe5887a","url":"https:\/\/secure.gravatar.com\/avatar\/6efa21323c55820fcce3e903fc80e9908b0b069289ee8cee40b589f840d8bbf8?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/6efa21323c55820fcce3e903fc80e9908b0b069289ee8cee40b589f840d8bbf8?s=96&d=mm&r=g","caption":"Chiara Binelli"},"url":"https:\/\/www.rivistaeco.com\/en\/author\/cbinelli\/"}]}},"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/posts\/4629","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/users\/7952"}],"replies":[{"embeddable":true,"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/comments?post=4629"}],"version-history":[{"count":1,"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/posts\/4629\/revisions"}],"predecessor-version":[{"id":4630,"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/posts\/4629\/revisions\/4630"}],"wp:attachment":[{"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/media?parent=4629"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/categories?post=4629"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/tags?post=4629"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.rivistaeco.com\/en\/wp-json\/wp\/v2\/coauthors?post=4629"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}