Scoring of tumor-infiltrating lymphocytes: From visual estimation to machine learning.

The extent of tumor-infiltrating lymphocytes (TILs), along with immunomodulatory ligands, tumor-mutational burden and other biomarkers, has been demonstrated to be a marker of response to immune-checkpoint therapy in several cancers. Pathologists have therefore started to devise standardized visual approaches to quantify TILs for therapy prediction. However, despite successful standardization efforts visual TIL estimation is slow, with limited precision and lacks the ability to evaluate more complex properties such as TIL distribution patterns. Therefore, computational image analysis approaches are needed to provide standardized and efficient TIL quantification. Here, we discuss different automated TIL scoring approaches ranging from classical image segmentation, where cell boundaries are identified and the resulting objects classified according to shape properties, to machine learning-based approaches that directly classify cells without segmentation but rely on large amounts of training data. In contrast to conventional machine learning (ML) approaches that are often criticized for their "black-box" characteristics, we also discuss explainable machine learning. Such approaches render ML results interpretable and explain the computational decision-making process through high-resolution heatmaps that highlight TILs and cancer cells and therefore allow for quantification and plausibility checks in biomedical research and diagnostics.

Display colour scale effects on diagnostic performance and reader agreement in cardiac CT and prostate apparent diffusion coefficient assessment.

AIM: To investigate the effect of colour scale choice on diagnostic performance in the interpretation of medical images. MATERIALS AND METHODS: Twelve clinicians interpreted 210 myocardial computed tomography (CT) perfusion (CTP) examinations, and nine clinicians interpreted 165 magnetic resonance imaging (MRI) apparent diffusion coefficient (ADC) prostate images. In three separate sessions, each participant read the same image set using greyscale, hot-iron, and rainbow scales, respectively. Participants scored their level of confidence for tumour presence in the ADC study, and for ischaemia in the CTP study, from 0 to 100. The area under the receiver operating characteristic (ROC) curve (AUC) was used as the performance metric. For cases that scored >50, CTP readers' agreement on the ischaemic transmural extent was analysed, and ADC map readers' selected values and coordinates for the lowest ADC within the detected tumour were compared across different colour scales. RESULTS: For CTP detection, the AUC was up to 0.10 higher with greyscale, 0.67±0.02 (standard error), compared to rainbow, 0.56±0.02, and detection with hot-iron was in between (0.61±0.03). For ischaemic transmural lesion categorisation, observed inter-reader agreement was highest with greyscale for category 25-50%. There is a small tendency for rainbow and greyscale to outperform hot-iron in the detection of prostate tumours. The selected lowest ADC value and pixel localisation was similar with all colour scales. CONCLUSIONS: The present findings suggest that colour visualisation has a measurable effect on CTP and ADC performance. Further investigation is necessary to determine the magnitude of the effect in diagnostic tasks.

Enriched designs for assessing discriminatory performance--analysis of bias and variance.

In evaluating discriminatory performance of a new modality in a screening setting, a logistical constraint is that the prevalence of the disease of interest is typically very low. This implies that under a standard study design large numbers of subjects have to be evaluated using the new modality. However, if a predicate modality exists in clinical practice, one can base inclusion into the study of the new modality on the clinical results from the predicate to 'enrich' the population of diseased subjects in the study. If this enrichment is not accounted for when estimating sensitivity, specificity, and area under the ROC curve, these 'naive' estimates may be substantially biased compared with expected performance in the intended use population. We derive expressions for the magnitude of this bias in terms of correlations of modality scores. When such estimates are 'corrected' for the sampling weights using inverse probability weighting, the variances of the estimates of the above quantities are affected. We derive here analytic expressions for these variances. For a fixed number of diseased subjects, differential sampling increases the variance of the (corrected) estimates, all other things being equal. However, differential sampling also increases the number with disease for fixed total study size, which decreases the variance of the sensitivity and area under the ROC curve estimates, all other things being equal. The balance of these two effects determines the gain in efficiency when using enrichment and corrected estimates. These principles are illustrated with a simulation study motivated by the Digital Mammographic Imaging Screening Trial study, a trial of digital versus screen film mammography.

Silver square nanospirals mimic optical properties of U-shaped metamaterials.

We present a study of the optical properties of three-armed square nanospirals made of silver and realized as nanostructured thin films with Glancing Angle Deposition. Calculation of current flows in the nanospirals show excited resonant modes resembling those observed in U-shaped resonators. Four principal resonances were determined: near 200 THz and 480 THz for one polarization and 250 THz and 650 THz for the polarization orthogonal to the first one. In particular, a mode with anti-parallel current flow in opposite arms, associated with the observed resonance near 650 THz, indicates the existence of a magnetic-like resonance in the square nanospiral arrays. The robustness of the resonances against variations in the structural parameters of the nanospirals was investigated. This study revealed that the main parameter driving the position of the resonances was the overall dimension of the nanospiral, directly related to the length of their arms. Optical properties of a sample were measured by generalized spectroscopic ellipsometry at near-normal incidence, and evidence conversion between polarization states even for light polarized in the plane containing one of the arms in agreement with the numerical study. The measurements compared favorably to the results of the numerical simulations taking into account the disorder in the sample.

Making an omnidirectional reflector.

The effect of having a finite number of layers on the design of omnidirectional reflectors was investigated. It was shown that the structure should be finished with a low-index layer having a thickness larger than a quarter-wave to increase reflectivity, whereas layers below may remain of quarter-wave optical thickness at normal incidence angle. This general trend has been used for designing and realizing two a-Si-SiO(2) (amorphous silicon and silicon dioxide) omnidirectional reflectors in the near-infrared range on a silicon and a silica substrate, respectively. Owing to the decrease of absorption of recrystallized silicon as compared with a-Si in the visible range, the transmissivity of the structure realized on silica substrate was dramatically increased in the visible range upon annealing, whereas the high reflectivity and the omnidirectional effect were maintained in the near-infrared range.