國立高雄大學圖資館 |

Language: English

Back

Explaining the Unexplainable: Medical Decision-Making, AI, and a Right To Explanation.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Explaining the Unexplainable: Medical Decision-Making, AI, and a Right To Explanation.
Author:	Lang, Michael Brian.
Published:	Ann Arbor : ProQuest Dissertations & Theses, 2022
Description:	129 p.
Notes:	Source: Masters Abstracts International, Volume: 84-05.
Notes:	Advisor: Zawati, Ma'n Hilmi;Beaudry, Jonas-Sebastien.
Contained By:	Masters Abstracts International84-05.
Subject:	Medical equipment.
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30157848
ISBN:	9798352989593

Explaining the Unexplainable: Medical Decision-Making, AI, and a Right To Explanation.
Lang, Michael Brian.

Explaining the Unexplainable: Medical Decision-Making, AI, and a Right To Explanation. - Ann Arbor : ProQuest Dissertations & Theses, 2022 - 129 p.

Source: Masters Abstracts International, Volume: 84-05.

Thesis (LL.M.)--McGill University (Canada), 2022.

This item must not be sold to any third party vendors.

Significant decisions in medicine are being increasingly delegated to machines. Automated machine learning models, many of which are thought to be at least as reliable and accurate as human decision-makers, are being used to make decisions about diagnosis, treatment, and care allocation. Though these systems will potentially enhance the quality of health outcomes and contribute to more efficient models of care delivery, they also pose an explanation challenge.Decisions made by machine learning models are often not accompanied by explanations: it is often technically impossible to know why a machine learning system reaches one decision rather than another. This raises difficult legal and ethical questions about responsibility, equality, and the fundamental principles of procedural law.This essay explores the degree to which unexplainable decision-making interferes with our conventional ways of understanding the practice and regulation of medicine. I suggest, using medical malpractice as a model, that the challenges posed by unexplainable machine learning may be profound. I describe how, in the face of unexplainable machine learning, several jurisdictions have enacted ‘rights to explanation,’ including Quebec and the European Union. But these emerging statutory rights are unlikely to respond adequately to the justice implications generated by unexplainable machine learning in medicine. In fact, rights to explanation will probably make things worse.

ISBN: 9798352989593Subjects--Topical Terms:

942653
Medical equipment.

Explaining the Unexplainable: Medical Decision-Making, AI, and a Right To Explanation.
LDR:04414nmm a2200385 4500 001 636208
005 20230501063928.5
006 m o d
007 cr#unu||||||||
008 230724s2022 ||||||||||||||||| ||eng d
020 $a 9798352989593
035 $a (MiAaPQ)AAI30157848
035 $a (MiAaPQ)McGill_hm50tx76b
035 $a AAI30157848
040 $a MiAaPQ $c MiAaPQ
100 1 $a Lang, Michael Brian. $3 942652
245 1 0 $a Explaining the Unexplainable: Medical Decision-Making, AI, and a Right To Explanation.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2022
300 $a 129 p.
500 $a Source: Masters Abstracts International, Volume: 84-05.
500 $a Advisor: Zawati, Ma'n Hilmi;Beaudry, Jonas-Sebastien.
502 $a Thesis (LL.M.)--McGill University (Canada), 2022.
506 $a This item must not be sold to any third party vendors.
520 $a Significant decisions in medicine are being increasingly delegated to machines. Automated machine learning models, many of which are thought to be at least as reliable and accurate as human decision-makers, are being used to make decisions about diagnosis, treatment, and care allocation. Though these systems will potentially enhance the quality of health outcomes and contribute to more efficient models of care delivery, they also pose an explanation challenge.Decisions made by machine learning models are often not accompanied by explanations: it is often technically impossible to know why a machine learning system reaches one decision rather than another. This raises difficult legal and ethical questions about responsibility, equality, and the fundamental principles of procedural law.This essay explores the degree to which unexplainable decision-making interferes with our conventional ways of understanding the practice and regulation of medicine. I suggest, using medical malpractice as a model, that the challenges posed by unexplainable machine learning may be profound. I describe how, in the face of unexplainable machine learning, several jurisdictions have enacted ‘rights to explanation,’ including Quebec and the European Union. But these emerging statutory rights are unlikely to respond adequately to the justice implications generated by unexplainable machine learning in medicine. In fact, rights to explanation will probably make things worse.
520 $a Les decisions importantes en medecine sont de plus en plus deleguees aux machines. Des modeles d’apprentissage machine automatises, dont beaucoup sont consideres comme au moins aussi fiables et precis que les decideurs humains, sont utilises pour prendre des decisions en matiere de diagnostic, de traitement et de repartition des soins. Bien que ces systemes soient susceptibles d’ameliorer la qualite des resultats en matiere de sante et de contribuer a des modeles plus efficaces de prestation de soins, ils posent egalement un probleme d'explication.Les decisions prises par les modeles d'apprentissage automatique ne sont souvent pas accompagnees d'explications: il est souvent techniquement impossible de savoir pourquoi un systeme d'apprentissage automatique parvient a une decision plutot qu'a une autre. Cela souleve des questions juridiques et ethiques difficiles sur la responsabilite, l’egalite et les principes fondamentaux du droit procedural.Cette these explore la mesure dans laquelle la prise de decision inexplicable interfere avec nos facons conventionnelles de comprendre la pratique et la reglementation de la medecine. Je suggere, en utilisant la faute medicale comme modele, que les defis poses par l’apprentissage automatique inexplicable sont susceptibles d’etre profonds. Je decris comment, face a l’apprentissage automatique inexplicable, plusieurs juridictions ont promulgue des ‘droits a l’explication,’ notamment le Quebec et l’Union europeenne. Mais il est peu probable que ces droits statutaires emergents repondent de maniere adequate aux implications de justice generees par l’apprentissage automatique inexplicable en medecine. En fait, les droits a l’explication sont susceptibles d’aggraver la situation.
590 $a School code: 0781.
650 4 $a Medical equipment. $3 942653
650 4 $a Medical malpractice. $3 942654
650 4 $a Decision making. $3 183849
650 4 $a Law. $3 207600
650 4 $a Medical ethics. $3 195041
650 4 $a Health care management. $3 730398
650 4 $a Medicine. $3 193819
690 $a 0800
690 $a 0398
690 $a 0497
690 $a 0769
690 $a 0564
710 2 $a McGill University (Canada). $3 531188
773 0 $t Masters Abstracts International $g 84-05.
790 $a 0781
791 $a LL.M.
792 $a 2022
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30157848