RESUMEN
Artificial intelligence (AI) algorithms have been retrospectively evaluated as replacement for one radiologist in screening mammography double-reading; however, methods for resolving discordance between radiologists and AI in the absence of 'real-world' arbitration may underestimate cancer detection rate (CDR) and recall. In 108,970 consecutive screens from a population screening program (BreastScreen WA, Western Australia), 20,120 were radiologist/AI discordant without real-world arbitration. Recall probabilities were randomly assigned for these screens in 1000 simulations. Recall thresholds for screen-detected and interval cancers (sensitivity) and no cancer (false-positive proportion, FPP) were varied to calculate mean CDR and recall rate for the entire cohort. Assuming 100% sensitivity, the maximum CDR was 7.30 per 1000 screens. To achieve >95% probability that the mean CDR exceeded the screening program CDR (6.97 per 1000), interval cancer sensitivities ≥63% (at 100% screen-detected sensitivity) and ≥91% (at 80% screen-detected sensitivity) were required. Mean recall rate was relatively constant across sensitivity assumptions, but varied by FPP. FPP > 6.5% resulted in recall rates that exceeded the program estimate (3.38%). CDR improvements depend on a majority of interval cancers being detected in radiologist/AI discordant screens. Such improvements are likely to increase recall, requiring careful monitoring where AI is deployed for screen-reading.
RESUMEN
BACKGROUND: Artificial intelligence (AI) has been proposed to reduce false-positive screens, increase cancer detection rates (CDRs), and address resourcing challenges faced by breast screening programs. We compared the accuracy of AI versus radiologists in real-world population breast cancer screening, and estimated potential impacts on CDR, recall and workload for simulated AI-radiologist reading. METHODS: External validation of a commercially-available AI algorithm in a retrospective cohort of 108,970 consecutive mammograms from a population-based screening program, with ascertained outcomes (including interval cancers by registry linkage). Area under the ROC curve (AUC), sensitivity and specificity for AI were compared with radiologists who interpreted the screens in practice. CDR and recall were estimated for simulated AI-radiologist reading (with arbitration) and compared with program metrics. FINDINGS: The AUC for AI was 0.83 compared with 0.93 for radiologists. At a prospective threshold, sensitivity for AI (0.67; 95% CI: 0.64-0.70) was comparable to radiologists (0.68; 95% CI: 0.66-0.71) with lower specificity (0.81 [95% CI: 0.81-0.81] versus 0.97 [95% CI: 0.97-0.97]). Recall rate for AI-radiologist reading (3.14%) was significantly lower than for the BSWA program (3.38%) (-0.25%; 95% CI: -0.31 to -0.18; P < 0.001). CDR was also lower (6.37 versus 6.97 per 1000) (-0.61; 95% CI: -0.77 to -0.44; P < 0.001); however, AI detected interval cancers that were not found by radiologists (0.72 per 1000; 95% CI: 0.57-0.90). AI-radiologist reading increased arbitration but decreased overall screen-reading volume by 41.4% (95% CI: 41.2-41.6). INTERPRETATION: Replacement of one radiologist by AI (with arbitration) resulted in lower recall and overall screen-reading volume. There was a small reduction in CDR for AI-radiologist reading. AI detected interval cases that were not identified by radiologists, suggesting potentially higher CDR if radiologists were unblinded to AI findings. These results indicate AI's potential role as a screen-reader of mammograms, but prospective trials are required to determine whether CDR could improve if AI detection was actioned in double-reading with arbitration. FUNDING: National Breast Cancer Foundation (NBCF), National Health and Medical Research Council (NHMRC).
Asunto(s)
Neoplasias de la Mama , Humanos , Femenino , Neoplasias de la Mama/diagnóstico por imagen , Neoplasias de la Mama/epidemiología , Inteligencia Artificial , Estudios Retrospectivos , Estudios Prospectivos , Estudios de Cohortes , Tamizaje Masivo/métodos , Detección Precoz del Cáncer/métodos , Mamografía/métodosRESUMEN
INTRODUCTION: Artiï¬cial intelligence (AI) algorithms for interpreting mammograms have the potential to improve the effectiveness of population breast cancer screening programmes if they can detect cancers, including interval cancers, without contributing substantially to overdiagnosis. Studies suggesting that AI has comparable or greater accuracy than radiologists commonly employ 'enriched' datasets in which cancer prevalence is higher than in population screening. Routine screening outcome metrics (cancer detection and recall rates) cannot be estimated from these datasets, and accuracy estimates may be subject to spectrum bias which limits generalisabilty to real-world screening. We aim to address these limitations by comparing the accuracy of AI and radiologists in a cohort of consecutive of women attending a real-world population breast cancer screening programme. METHODS AND ANALYSIS: A retrospective, consecutive cohort of digital mammography screens from 109 000 distinct women was assembled from BreastScreen WA (BSWA), Western Australia's biennial population screening programme, from November 2016 to December 2017. The cohort includes 761 screen-detected and 235 interval cancers. Descriptive characteristics and results of radiologist double-reading will be extracted from BSWA outcomes data collection. Mammograms will be reinterpreted by a commercial AI algorithm (DeepHealth). AI accuracy will be compared with that of radiologist single-reading based on the diï¬erence in the area under the receiver operating characteristic curve. Cancer detection and recall rates for combined AI-radiologist reading will be estimated by pairing the first radiologist read per screen with the AI algorithm, and compared with estimates for radiologist double-reading. ETHICS AND DISSEMINATION: This study has ethical approval from the Women and Newborn Health Service Ethics Committee (EC00350) and the Curtin University Human Research Ethics Committee (HRE2020-0316). Findings will be published in peer-reviewed journals and presented at national and international conferences. Results will also be disseminated to stakeholders in Australian breast cancer screening programmes and policy makers in population screening.
Asunto(s)
Neoplasias de la Mama , Detección Precoz del Cáncer , Inteligencia Artificial , Australia , Neoplasias de la Mama/diagnóstico por imagen , Estudios de Cohortes , Detección Precoz del Cáncer/métodos , Femenino , Humanos , Recién Nacido , Mamografía/métodos , Tamizaje Masivo , Estudios RetrospectivosRESUMEN
Autosomal recessive mutations in the PINK1 gene are causal for Parkinson's disease (PD). PINK1 encodes a mitochondrial localized protein kinase that is a master-regulator of mitochondrial quality control pathways. Structural studies to date have elaborated the mechanism of how mutations located within the kinase domain disrupt PINK1 function; however, the molecular mechanism of PINK1 mutations located upstream and downstream of the kinase domain is unknown. We have employed mutagenesis studies to define the minimal region of human PINK1 required for optimal ubiquitin phosphorylation, beginning at residue Ile111. Inspection of the AlphaFold human PINK1 structure model predicts a conserved N-terminal α-helical extension (NTE) domain forming an intramolecular interaction with the C-terminal extension (CTE), which we corroborate using hydrogen/deuterium exchange mass spectrometry of recombinant insect PINK1 protein. Cell-based analysis of human PINK1 reveals that PD-associated mutations (e.g. Q126P), located within the NTE : CTE interface, markedly inhibit stabilization of PINK1; autophosphorylation at Serine228 (Ser228) and Ubiquitin Serine65 (Ser65) phosphorylation. Furthermore, we provide evidence that NTE and CTE domain mutants disrupt PINK1 stabilization at the mitochondrial Translocase of outer membrane complex. The clinical relevance of our findings is supported by the demonstration of defective stabilization and activation of endogenous PINK1 in human fibroblasts of a patient with early-onset PD due to homozygous PINK1 Q126P mutations. Overall, we define a functional role of the NTE : CTE interface towards PINK1 stabilization and activation and show that loss of NTE : CTE interactions is a major mechanism of PINK1-associated mutations linked to PD.
Asunto(s)
Proteínas Quinasas , Ubiquitina , Activación Enzimática , Humanos , Fosforilación , Conformación Proteica en Hélice alfa , Proteínas Quinasas/genética , Proteínas Quinasas/metabolismo , Ubiquitina/metabolismo , Ubiquitina-Proteína Ligasas/metabolismoRESUMEN
Mutations in the human kinase PINK1 (hPINK1) are associated with autosomal recessive early-onset Parkinson's disease (PD). hPINK1 activates Parkin E3 ligase activity, involving phosphorylation of ubiquitin and the Parkin ubiquitin-like (Ubl) domain via as yet poorly understood mechanisms. hPINK1 is unusual amongst kinases due to the presence of three loop insertions of unknown function. We report the structure of Tribolium castaneum PINK1 (TcPINK1), revealing several unique extensions to the canonical protein kinase fold. The third insertion, together with autophosphorylation at residue Ser205, contributes to formation of a bowl-shaped binding site for ubiquitin. We also define a novel structural element within the second insertion that is held together by a distal loop that is critical for TcPINK1 activity. The structure of TcPINK1 explains how PD-linked mutations that lie within the kinase domain result in hPINK1 loss-of-function and provides a platform for the exploration of small molecule modulators of hPINK1.