AI in Pathology: What could possibly go wrong?

There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

Abstract

<p xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" class="first" dir="auto" id="d5841926e158">The field of medicine is undergoing rapid digital transformation. Pathologists are now striving to digitize their data, workflows, and interpretations, assisted by the enabling development of whole-slide imaging. Going digital means that the analog process of human diagnosis can be augmented or even replaced by rapidly evolving AI approaches, which are just now entering into clinical practice. But with such progress comes challenges that reflect a variety of stressors, including the impact of unrepresentative training data with accompanying implicit bias, data privacy concerns, and fragility of algorithm performance. Beyond such core digital aspects, considerations arise related to difficulties presented by changing disease presentations, diagnostic approaches, and therapeutic options. While some tools such as data federation can help with broadening data diversity while preserving expertise and local control, they may not be the full answer to some of these issues. The impact of AI in pathology on the field's human practitioners is still very much unknown: installation of unconscious bias and deference to AI guidance need to be understood and addressed. If AI is widely adopted, it may remove many inefficiencies in daily practice and compensate for staff shortages. It may also cause practitioner deskilling, dethrilling, and burnout. We discuss the technological, clinical, legal, and sociological factors that will influence the adoption of AI in pathology, and its eventual impact for good or ill. </p>

Related collections

Most cited references 60

Record: found
Abstract: found
Article: not found

Burnout and satisfaction with work-life balance among US physicians relative to the general US population.

Tait Shanafelt, Sonja Boone, Litjen Tan … (2012)

Despite extensive data about physician burnout, to our knowledge, no national study has evaluated rates of burnout among US physicians, explored differences by specialty, or compared physicians with US workers in other fields. We conducted a national study of burnout in a large sample of US physicians from all specialty disciplines using the American Medical Association Physician Masterfile and surveyed a probability-based sample of the general US population for comparison. Burnout was measured using validated instruments. Satisfaction with work-life balance was explored. Of 27 276 physicians who received an invitation to participate, 7288 (26.7%) completed surveys. When assessed using the Maslach Burnout Inventory, 45.8% of physicians reported at least 1 symptom of burnout. Substantial differences in burnout were observed by specialty, with the highest rates among physicians at the front line of care access (family medicine, general internal medicine, and emergency medicine). Compared with a probability-based sample of 3442 working US adults, physicians were more likely to have symptoms of burnout (37.9% vs 27.8%) and to be dissatisfied with work-life balance (40.2% vs 23.2%) (P < .001 for both). Highest level of education completed also related to burnout in a pooled multivariate analysis adjusted for age, sex, relationship status, and hours worked per week. Compared with high school graduates, individuals with an MD or DO degree were at increased risk for burnout (odds ratio [OR], 1.36; P < .001), whereas individuals with a bachelor's degree (OR, 0.80; P = .048), master's degree (OR, 0.71; P = .01), or professional or doctoral degree other than an MD or DO degree (OR, 0.64; P = .04) were at lower risk for burnout. Burnout is more common among physicians than among other US workers. Physicians in specialties at the front line of care access seem to be at greatest risk.

0 comments Cited 507 times – based on 0 reviews      Review now

Bookmark

Record: found
Abstract: found
Article: found

Is Open Access

The 5th edition of the World Health Organization Classification of Haematolymphoid Tumours: Lymphoid Neoplasms

Rita Alaggio, Catalina Amador, Ioannis Anagnostopoulos … (2022)

We herein present an overview of the upcoming 5th edition of the World Health Organization Classification of Haematolymphoid Tumours focussing on lymphoid neoplasms. Myeloid and histiocytic neoplasms will be presented in a separate accompanying article. Besides listing the entities of the classification, we highlight and explain changes from the revised 4th edition. These include reorganization of entities by a hierarchical system as is adopted throughout the 5th edition of the WHO classification of tumours of all organ systems, modification of nomenclature for some entities, revision of diagnostic criteria or subtypes, deletion of certain entities, and introduction of new entities, as well as inclusion of tumour-like lesions, mesenchymal lesions specific to lymph node and spleen, and germline predisposition syndromes associated with the lymphoid neoplasms.

0 comments Cited 440 times – based on 0 reviews      Review now

Bookmark

Record: found
Abstract: found
Article: not found

Artificial intelligence in digital pathology — new tools for diagnosis and precision oncology

Kaustav Bera, Kurt Schalper, David Rimm … (2019)

In the past decade, advances in precision oncology have resulted in an increased demand for predictive assays that enable the selection and stratification of patients for treatment. The enormous divergence of signalling and transcriptional networks mediating the crosstalk between cancer, stromal and immune cells complicates the development of functionally relevant biomarkers based on a single gene or protein. However, the result of these complex processes can be uniquely captured in the morphometric features of stained tissue specimens. The possibility of digitizing whole-slide images of tissue has led to the advent of artificial intelligence (AI) and machine learning tools in digital pathology, which enable mining of subvisual morphometric phenotypes and might, ultimately, improve patient management. In this Perspective, we critically evaluate various Al-based computational approaches for digital pathology, focusing on deep neural networks and ‘hand-crafted’ feature-based methodologies. We aim to provide a broad framework for incorporating AI and machine learning tools into clinical oncology, with an emphasis on biomarker development. We discuss some of the challenges relating to the use of AI, including the need for well-curated validation datasets, regulatory approval and fair reimbursement strategies. Finally, we present potential future opportunities for precision oncology.