RESUMEN
INTRODUCTION: In the last decade, various Machine Learning techniques have been proposed aiming to individualise the dose of anticancer drugs mostly based on a presumed drug effect or measured effect biomarkers. The aim of this scoping review was to comprehensively summarise the research status on the use of Machine Learning for precision dosing in anticancer drug therapy. METHODS: This scoping review was conducted in accordance with the interim guidance by Cochrane and the Joanna Briggs Institute. We systematically searched the databases Medline (via PubMed), Embase and the Cochrane Library for research articles and reviews including results published after 2016. Results were reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist. RESULTS: A total of 17 relevant studies was identified. In 12 of the included studies, Reinforcement Learning methods were used, including Classical, Deep, Double Deep and Conservative Q-Learning and Fuzzy Reinforcement Learning. Furthermore, classical Machine Learning methods were compared in terms of their performance and an artificial intelligence platform based on parabolic equations was used to guide dosing prospectively and retrospectively, albeit only in a limited number of patients. Due to the significantly different algorithm structures, a meaningful comparison between the various Machine Learning approaches was not possible. CONCLUSION: Overall, this review emphasises the clinical relevance of Machine Learning methods for anticancer drug dose optimisation, as many algorithms have shown promising results enabling model-free predictions with the potential to maximise efficacy and minimise toxicity when compared to standard protocols.
RESUMEN
Background: Drusen and drusenoid deposits are a hallmark of age-related macular degeneration (AMD). Nowadays, a multimodal retinal imaging approach enables the detection of these deposits. However, quantitative data on subretinal drusenoid deposits (SDDs) are still missing. Here, we compare the capability of en-face drusen and SDD area detection in eyes with non-exudative AMD using conventional imaging modalities versus Retro mode imaging. We also quantitatively assess the topographic distribution of drusen and SDDs. Methods: In total, 120 eyes of 90 subjects (mean age ± standard deviation = 74.6 ± 8.6 years) were included. Coherent en-face drusen and SDD areas were measured via near-infrared reflectance, green (G-) and blue (B-) fundus autofluorescence (AF), and Retro mode imaging. Drusen phenotypes were classified by correlating en-face drusen areas using structural high-resolution spectral domain optical coherence tomography. The topographic distribution of drusen was analyzed according to a modified ETDRS (Early Treatment of Diabetic Retinopathy Study) grid. Intraclass correlation coefficient (ICC) analysis was applied to determine the inter-reader agreement in the SDD en-face area assessment. Results: The largest coherent en-face drusen area was found using Retro mode imaging with a mean area of 105.2 ± 45.9 mm2 (deviated left mode (DL)) and 105.4 ± 45.5 mm2 (deviated right mode (DR)). The smallest en-face drusen areas were determined by GAF (50.9 ± 42.6 mm2) and BAF imaging (49.1 ± 42.9 mm2) (p < 0.001). The inter-reader agreement for SDD en-face areas ranged from 0.93 (DR) to 0.70 (BAF). The topographic analysis revealed the highest number of SDDs in the superior peripheral retina, whereas sub-retinal pigment epithelium drusen were mostly found in the perifoveal retina. Retro mode imaging further enabled the detection of the earliest SDD stages. Conclusions: Retro mode imaging allows for a detailed detection of drusen phenotypes. While hundreds/thousands of SDDs can be present in one eye, the impact of SDD number or volume on AMD progression still needs to be evaluated. However, this new imaging modality can add important knowledge on drusen development and the pathophysiology of AMD.
RESUMEN
Purpose: To longitudinally assess the impact of high-risk structural biomarkers for natural disease progression in non-exudative age-related macular degeneration (AMD) on spatially resolved mesopic and scotopic fundus-controlled perimetry testing. Methods: Multimodal retinal imaging data and fundus-controlled perimetry stimuli points were semiautomatically registered according to landmark correspondences at each annual visit over a period of up to 4 years. The presence of sub-RPE drusen, subretinal drusenoid deposits, pigment epithelium detachments (PEDs), hyper-reflective foci (HRF), vitelliform lesions, refractile deposits, and incomplete RPE and outer retinal atrophy (iRORA) and complete RPE and outer retinal atrophy (cRORA) were graded at each stimulus position and visit. Localized retinal layer thicknesses were extracted. Mixed-effect models were used for structure-function correlation. Results: Fifty-four eyes of 49 patients with non-exudative AMD (mean age, 70.7 ± 9.1 years) and 27 eyes of 27 healthy controls (mean age, 63.4 ± 8.9 years) were included. During study course, presence of PED had the highest functional impact with a mean estimated loss of -1.30 dB (P < 0.001) for mesopic and -1.23 dB (P < 0.001) for scotopic testing, followed by HRF with -0.89 dB (mesopic, P = 0.001) and -0.87 dB (scotopic, P = 0.005). Subretinal drusenoid deposits were associated with a stronger visual impairment (mesopic, -0.38 dB; P = 0.128; scotopic, -0.37 dB; P = 0.172) compared with sub-RPE drusen (-0.22 dB, P = 0.0004; -0.18 dB, P = 0.006). With development of c-RORA, scotopic retinal sensitivity further significantly decreased (-2.15 dB; P = 0.02). Thickening of the RPE-drusen-complex and thinning of the outer nuclear layer negatively impacted spatially resolved retinal sensitivity. Conclusions: The presence of PED and HRF had the greatest prognostic impact on progressive point-wise sensitivity losses. Higher predominant rod than cone-mediated localized retinal sensitivity losses with early signs of retinal atrophy development indicate photoreceptor preservation as a potential therapeutic target for future interventional AMD trials.
Asunto(s)
Progresión de la Enfermedad , Tomografía de Coherencia Óptica , Agudeza Visual , Pruebas del Campo Visual , Campos Visuales , Humanos , Femenino , Anciano , Masculino , Persona de Mediana Edad , Tomografía de Coherencia Óptica/métodos , Agudeza Visual/fisiología , Campos Visuales/fisiología , Degeneración Macular/fisiopatología , Degeneración Macular/diagnóstico , Drusas Retinianas/fisiopatología , Drusas Retinianas/diagnóstico , Biomarcadores , Estudios de Seguimiento , Epitelio Pigmentado de la Retina/patología , Epitelio Pigmentado de la Retina/fisiopatología , Visión Nocturna/fisiología , Retina/fisiopatología , Retina/diagnóstico por imagen , Retina/patología , Anciano de 80 o más Años , Angiografía con Fluoresceína/métodosRESUMEN
BACKGROUND: Retinal imaging, including fundus autofluorescence (FAF), strongly depends on the clearness of the optical media. Lens status is crucial since the ageing lens has both light-blocking and autofluorescence (AF) properties that distort image analysis. Here, we report both lens opacification and AF metrics and the effect on automated image quality assessment. METHODS: 227 subjects (range: 19-89 years old) received quantitative AF of the lens (LQAF), Scheimpflug, anterior chamber optical coherence tomography as well as blue/green FAF (BAF/GAF), and infrared (IR) imaging. LQAF values, the Pentacam Nucleus Staging score and the relative lens reflectivity were extracted to estimate lens opacification. Mean opinion scores of FAF and IR image quality were compiled by medical readers. A regression model for predicting image quality was developed using a convolutional neural network (CNN). Correlation analysis was conducted to assess the association of lens scores, with retinal image quality derived from human or CNN annotations. RESULTS: Retinal image quality was generally high across all imaging modalities (IR (8.25±1.99) >GAF >BAF (6.6±3.13)). CNN image quality prediction was excellent (average mean absolute error (MAE) 0.9). Predictions were comparable to human grading. Overall, LQAF showed the highest correlation with image quality grading criteria for all imaging modalities (eg, Pearson correlation±CI -0.35 (-0.50 to 0.18) for BAF/LQAF). BAF image quality was most vulnerable to an increase in lenticular metrics, while IR (-0.19 (-0.38 to 0.01)) demonstrated the highest resilience. CONCLUSION: The use of CNN-based retinal image quality assessment achieved excellent results. The study highlights the vulnerability of BAF to lenticular remodelling. These results can aid in the development of cut-off values for clinical studies, ensuring reliable data collection for the monitoring of retinal diseases.
Asunto(s)
Catarata , Tomografía de Coherencia Óptica , Humanos , Anciano , Persona de Mediana Edad , Adulto , Anciano de 80 o más Años , Femenino , Masculino , Tomografía de Coherencia Óptica/métodos , Adulto Joven , Catarata/diagnóstico por imagen , Catarata/patología , Retina/diagnóstico por imagen , Retina/patología , Imagen Óptica/métodos , Cristalino/diagnóstico por imagen , Cristalino/patología , Angiografía con Fluoresceína/métodosRESUMEN
Purpose: Quantitative fundus autofluorescence (QAF) currently deploys an age-based score to correct for lens opacification. However, in elderly people, lens opacification varies strongly between individuals of similar age, and innate lens autofluorescence is not included in the current correction formula. Our goal was to develop and compare an individualized formula. Methods: One hundred thirty participants were examined cross-sectionally, and a subset of 30 participants received additional multimodal imaging 2-week post-cataract-surgery. Imaging included the Scheimpflug principle, anterior chamber optical coherence tomography (AC-OCT), lens quantitative autofluorescence (LQAF), and retinal QAF imaging. Among the subset, least absolute shrinkage and selection operator regression and backward selection was implemented to determine which lens score best predicts the QAF value after lens extraction. Subsequently, a spline mixed model was applied to the whole cohort to quantify the influence of LQAF and Scheimpflug on QAF. Results: Age and LQAF measurements were found to be the most relevant variables, whereas AC-OCT measurements and Scheimpflug were eliminated by backward selection. Both an increase in Scheimpflug and LQAF values were associated with a decrease in QAF. The prediction error of the spline model (mean absolute error [MAE] ± standard deviation) of 32.2 ± 23.4 (QAF a.u.) was markedly lower compared to the current age-based formula MAE of 96.1 ± 93.5. Both smooth terms, LQAF (P < 0.01) and Scheimpflug (P < 0.001), were significant for the spline mixed model. Conclusions: LQAF imaging proved to be the most predictive for the impact of the natural lens on QAF imaging. The application of lens scores in the clinic could improve the accuracy of QAF imaging interpretation and might allow including aged patients in future QAF studies.
Asunto(s)
Extracción de Catarata , Catarata , Cristalino , Anciano , Humanos , Cristalino/diagnóstico por imagen , Fondo de Ojo , RetinaRESUMEN
Purpose: Retinal pigment epithelium (RPE) cells show strong autofluorescence (AF). Here, we characterize the AF spectra of individual RPE cells in healthy eyes and those affected by age-related macular degeneration (AMD) and investigate associations between AF spectral response and the number of intracellular AF granules per cell. Methods: RPE-Bruch's membrane flatmounts of 22 human donor eyes, including seven AMD-affected eyes (early AMD, three; geographic atrophy, one; neovascular, three) and 15 unaffected macula (<51 years, eight; >80 years, seven), were imaged at the fovea, perifovea, and near-periphery using confocal AF microscopy (excitation 488 nm), and emission spectra were recorded (500-710 nm). RPE cells were manually segmented with computer assistance and stratified by disease status, and emission spectra were analyzed using cubic spline transforms. Intracellular granules were manually counted and classified. Linear mixed models were used to investigate associations between spectra and the number of intracellular granules. Results: Spectra of 5549 RPE cells were recorded. The spectra of RPE cells in healthy eyes showed similar emission curves that peaked at 580 nm for fovea and perifovea and at 575 and 580 nm for near-periphery. RPE spectral curves in AMD eyes differed significantly, being blue shifted by 10 nm toward shorter wavelengths. No significant association coefficients were found between wavelengths and granule counts. Conclusions: This large series of RPE cell emission spectra at precisely predefined retinal locations showed a hypsochromic spectral shift in AMD. Combining different microscopy techniques, our work has identified cellular RPE spectral AF and subcellular granule properties that will inform future in vivo investigations using single-cell imaging.
Asunto(s)
Atrofia Geográfica , Mácula Lútea , Degeneración Macular , Humanos , Epitelio Pigmentado de la Retina/metabolismo , Degeneración Macular/diagnóstico , Degeneración Macular/metabolismo , Lámina Basal de la Coroides/metabolismo , Atrofia Geográfica/metabolismo , Mácula Lútea/metabolismoRESUMEN
This study aimed to determine the retest variability of quantitative fundus autofluorescence (QAF) in patients with and without age-related macular degeneration (AMD) and evaluate the predictive value of patient reliability indices on retest reliability. A total of 132 eyes from 68 patients were examined, including healthy individuals and those with various stages of AMD. Duplicate QAF imaging was conducted at baseline and 2 weeks later across six study sites. Intraclass correlation (ICC) analysis was used to evaluate the consistency of imaging, and mean opinion scores (MOS) of image quality were generated by two researchers. The contribution of MOS and other factors to retest variation was assessed using mixed-effect linear models. Additionally, a Random Forest Regressor was trained to evaluate the extent to which manual image grading of image quality could be replaced by automated assessment (inferred MOS). The results showed that ICC values were high for all QAF images, with slightly lower values in AMD-affected eyes. The average inter-day ICC was found to be 0.77 for QAF segments within the QAF8 ring and 0.74 for peripheral segments. Image quality was predicted with a mean absolute error of 0.27 on a 5-point scale, and of all evaluated reliability indices, MOS/inferred MOS proved most important. The findings suggest that QAF allows for reliable testing of autofluorescence levels at the posterior pole in patients with AMD in a multicenter, multioperator setting. Patient reliability indices could serve as eligibility criteria for clinical trials, helping identify patients with adequate retest reliability.
Asunto(s)
Degeneración Macular , Epitelio Pigmentado de la Retina , Humanos , Reproducibilidad de los Resultados , Angiografía con Fluoresceína/métodos , Fondo de Ojo , Degeneración Macular/diagnóstico por imagenRESUMEN
The current gold standard of response assessment in patients with myelodysplastic syndromes (MDS), chronic myelomonocytic leukemia (CMML), and acute myeloid leukemia (AML) is morphologic complete remission (CR) and CR with incomplete count recovery (CRi), both of which require an invasive BM evaluation. Outside of clinical trials, BM evaluations are only performed in ~50% of patients during follow-up, pinpointing a clinical need for response endpoints that do not necessitate BM assessments. We define and validate a new response type termed "peripheral blood complete remission" (PB-CR) that can be determined from the differential blood count and clinical parameters without necessitating a BM assessment. We compared the predictive value of PB-CR with morphologic CR/CRi in 1441 non-selected, consecutive patients diagnosed with MDS (n = 522; 36.2%), CMML (n = 132; 9.2%), or AML (n = 787; 54.6%), included within the Austrian Myeloid Registry (aMYELOIDr; NCT04438889). Time-to-event analyses were adjusted for 17 covariates remaining in the final Cox proportional hazards (CPH) model. DeepSurv, a CPH neural network model, and permutation-based feature importance were used to validate results. 1441 patients were included. Adjusted median overall survival for patients achieving PB-CR was 22.8 months (95%CI 18.9-26.2) versus 10.4 months (95%CI 9.7-11.2) for those who did not; HR = 0.366 (95%CI 0.303-0.441; p < .0001). Among patients achieving CR, those additionally achieving PB-CR had a median adjusted OS of 32.6 months (95%CI 26.2-49.2) versus 21.7 months (95%CI 16.9-27.7; HR = 0.400 [95%CI 0.190-0.844; p = .0161]) for those who did not. Our deep neural network analysis-based findings from a large, prospective cohort study indicate that BM evaluations solely for the purpose of identifying CR/CRi can be omitted.
RESUMEN
Accurate somatic variant calling from next-generation sequencing data is one most important tasks in personalised cancer therapy. The sophistication of the available technologies is ever-increasing, yet, manual candidate refinement is still a necessary step in state-of-the-art processing pipelines. This limits reproducibility and introduces a bottleneck with respect to scalability. We demonstrate that the validation of genetic variants can be improved using a machine learning approach resting on a Convolutional Neural Network, trained using existing human annotation. In contrast to existing approaches, we introduce a way in which contextual data from sequencing tracks can be included into the automated assessment. A rigorous evaluation shows that the resulting model is robust and performs on par with trained researchers following published standard operating procedure.
Asunto(s)
Aprendizaje Automático , Redes Neurales de la Computación , Humanos , Reproducibilidad de los Resultados , Secuenciación de Nucleótidos de Alto Rendimiento/métodosRESUMEN
In this prospective study (NCT01595295), 272 patients treated with azacitidine completed 1456 EuroQol 5-Dimension (EQ-5D) questionnaires. Linear mixed-effect modelling was used to incorporate longitudinal data. When compared with a matched reference population, myeloid patients reported more pronounced restrictions in usual activities (+28%, p < 0.0001), anxiety/depression (+21%, p < 0.0001), selfcare (+18%, p < 0.0001) and mobility (+15%, p < 0.0001), as well as lower mean EQ-5D-5L indices (0.81 vs. 0.88, p < 0.0001), and lower self-rated health on the EuroQol Visual Analogue Scale (EQ-VAS) (64 vs. 72%, p < 0.0001). After multivariate-adjustment, (i) the EQ-5D-5L index assessed at azacitidine start the predicted time with clinical benefit (TCB) (9.6 vs. 6.6 months; p = 0.0258; HR = 1.43), time to next treatment (TTNT) (12.8 vs. 9.8 months; p = 0.0332; HR = 1.42) and overall survival (OS) (17.9 vs. 12.9 months; p = 0.0143; HR = 1.52); (ii) Level Sum Score (LSS) predicted azacitidine response (p = 0.0160; OR = 0.451) and the EQ-5D-5L index showed a trend (p = 0.0627; OR = 0.522); (iii) up to 1432 longitudinally assessed EQ-5D-5L response/clinical parameter pairs revealed significant associations of EQ-5D-5L response parameters with haemoglobin level, transfusion dependence and hematologic improvement. Significant increases of the likelihood ratios were observed after addition of LSS, EQ-VAS or EQ-5D-5L-index to the International Prognostic Scoring System (IPSS) or the revised IPSS (R-IPSS), indicating that they provide added value to these scores.
RESUMEN
Purpose: The purpose of this study was to analyze spatially resolved structural changes at retinal locations in presence (+) or absence (-) of hyper-reflective foci (HRF) in eyes with subretinal pigment epithelium (RPE) drusen in intermediate age-related macular degeneration (iAMD). Methods: Patients with IAMD (n = 40; mean age = 69.7 ± 9.2 [SD] years) and healthy controls (n = 27; 64.2 ± 9.0) underwent spectral-domain optical-coherence-tomography imaging and fundus-controlled perimetry testing. After reviewing retinal layer segmentation, presence of HRF was annotated and retinal layer thicknesses (RLTs) extracted using ImageJ. Localized RLTs were compared between +HRF and -HRF positions. Univariate mixed linear models were used to investigate associations among RLT, HRF presence, and HRF size. Results: In iAMD eyes, a mean of 11.1 ± 12.5 HRF were detected with a peak abundance at 0.5 to 1.5 mm eccentricity to the fovea. At +HRF positions, outer nuclear layer (ONL; P = 0.0013, average difference = -12.4 µm) and retinal pigment epithelium drusen complex (RPEDC; P < 0.0001, +45.6 µm) thicknesses differed significantly compared to -HRF positions, even after correcting for accompanying drusen-related RPEDC layer thickening (P = 0.01). Mixed linear models revealed a significant association between increasing HRF area and decreasing ONL (association score = -0.17, P < 0.0001; 95% confidence interval [CI] = -0.22 to -0.11), and inner photoreceptor segments (IS) layer thicknesses (-0.08, P = 0.005; 95% CI = -0.14 to -0.03). Spearman rank correlation analysis yielded a significant correlation between total HRF area and mesopic (P = 0.015), but not scotopic (P = 0.305) retinal sensitivity losses. Conclusions: Descriptive analysis of this study demonstrated a predominant distribution of HRF at a foveal eccentricity of 0.5 to 1.5 mm, whereas further refined topographic analysis revealed a significant ONL layer thinning in presence of HRF even after correction for sub-RPE drusen presence compared to lesions in absence of HRF. Longitudinal studies are further needed to analyze the prognostic impact as well as the role of HRF presence in the context of iAMD.
Asunto(s)
Degeneración Macular , Drusas Retinianas , Humanos , Persona de Mediana Edad , Anciano , Drusas Retinianas/diagnóstico , Drusas Retinianas/patología , Retina/diagnóstico por imagen , Retina/patología , Degeneración Macular/diagnóstico , Degeneración Macular/patología , Epitelio Pigmentado de la Retina/patología , Pruebas del Campo Visual , Tomografía de Coherencia Óptica/métodosRESUMEN
Purpose: Phenotype alterations of the retinal pigment epithelium (RPE) are a main characteristic of age-related macular degeneration (AMD). Individual RPE cell shape descriptors may help to delineate healthy from AMD-affected cells in early disease stages. Methods: Twenty-two human RPE flatmounts (7 eyes with AMD [early, 3; geographic atrophy, 1; neovascular, 3); 15 unaffected eyes [8 aged ≤51 years; 7 aged >80 years)] were imaged at the fovea, perifovea, and near periphery (predefined sample locations) using a laser-scanning confocal fluorescence microscope. RPE cell boundaries were manually marked with computer assistance. For each cell, 11 shape descriptors were calculated and correlated with donor age, cell autofluorescence (AF) intensity, and retinal location. Statistical analysis was performed using an ensemble classifier based on logistic regression. Results: In AMD, RPE was altered at all locations (most pronounced at the fovea), with area, solidity, and form factor being the most discriminatory descriptors. In the unaffected macula, aging had no significant effect on cell shape factors; however, with increasing distance to the fovea, area, solidity, and convexity increased while form factor decreased. Reduced AF in AMD was significantly associated with decreased roundness and solidity. Conclusions: AMD results in an altered RPE with enlarged and deformed cells that could precede clinically visible lesions and thus serve as early biomarkers for AMD onset. Our data may also help guide the interpretation of RPE morphology in in vivo studies utilizing high-resolution single-cell imaging. Translational Relevance: Our histologic RPE cell shape data have the ability to identify robust biomarkers for the early detection of AMD-affected cells, which also could serve as a basis for automated segmentation of RPE sheets.