Explaining neural scaling laws.

The population loss of trained deep neural networks often follows precise power-law scaling relations with either the size of the training dataset or the number of parameters in the network. We propose a theory that explains the origins of and connects these scaling laws. We identify variance-limited and resolution-limited scaling behavior for both dataset and model size, for a total of four scaling regimes. The variance-limited scaling follows simply from the existence of a well-behaved infinite data or infinite width limit, while the resolution-limited regime can be explained by positing that models are effectively resolving a smooth data manifold. In the large width limit, this can be equivalently obtained from the spectrum of certain kernels, and we present evidence that large width and large dataset resolution-limited scaling exponents are related by a duality. We exhibit all four scaling regimes in the controlled setting of large random feature and pretrained models and test the predictions empirically on a range of standard architectures and datasets. We also observe several empirical relationships between datasets and scaling exponents under modifications of task and architecture aspect ratio. Our work provides a taxonomy for classifying different scaling regimes, underscores that there can be different mechanisms driving improvements in loss, and lends insight into the microscopic origin and relationships between scaling exponents.

Early observations with an ERAS pathway for thyroid and parathyroid surgery: Moving the goalposts forward.

BACKGROUND: Enhanced recovery after surgery pathways have become the standard of care in various surgical specialties. In this study, we discuss our initial experience with a staged enhanced recovery after surgery pathway in endocrine surgery and assess the impact of this pathway on select perioperative outcomes and unanticipated admissions. METHODS: We collected information regarding all thyroid/parathyroid surgeries performed by endocrine surgeons at our institution before and after the implementation of the multi-intervention enhanced recovery after surgery pathway. We compared relevant outcomes for all cases 1 year before (n = 479) and 1 year after (n = 166) implementation of the pathway. We also compared outcomes between enhanced recovery after surgery patient groups with varying levels of enhanced recovery after surgery compliance. RESULTS: Enhanced recovery after surgery was associated with a significant decrease in total length of stay (9.2 vs 7.5 hours, P < .0001). Whereas there was no significant decrease in all-cause unanticipated postoperative admissions, there was a decrease in patient-initiated admissions in the Enhanced recovery after surgery group. There was also a significant decrease in mean postoperative morphine milligram equivalents (14.4 vs 16.2 vs 24.8, P = .0015), average daily morphine milligram equivalents (25.6 vs 45.6 vs 53, P < .0001), and average daily pain scores (1.89 vs 2.38 vs 2.74, P = .0045) in the Enhanced recovery after surgery group (particularly with increasing Enhanced recovery after surgery compliance). There were no significant differences in the requirement for postoperative antiemetics or in the post-anesthesia care unit length of stay. CONCLUSION: This study demonstrates a significant benefit from Enhanced recovery after surgery pathways for thyroidectomies and parathyroidectomies, even with initial data and a staggered roll-out plan. Further directions include a follow-up study once we reach a higher level of institutional compliance with all components of the Enhanced Recovery After Surgery pathway and a prospective trial to identify the relative significance of different portions of the Enhanced Recovery after Surgery pathway, particularly the superficial cervical plexus block.

WhichTF is functionally important in your open chromatin data?

We present WhichTF, a computational method to identify functionally important transcription factors (TFs) from chromatin accessibility measurements. To rank TFs, WhichTF applies an ontology-guided functional approach to compute novel enrichment by integrating accessibility measurements, high-confidence pre-computed conservation-aware TF binding sites, and putative gene-regulatory models. Comparison with prior sheer abundance-based methods reveals the unique ability of WhichTF to identify context-specific TFs with functional relevance, including NF-κB family members in lymphocytes and GATA factors in cardiac cells. To distinguish the transcriptional regulatory landscape in closely related samples, we apply differential analysis and demonstrate its utility in lymphocyte, mesoderm developmental, and disease cells. We find suggestive, under-characterized TFs, such as RUNX3 in mesoderm development and GLI1 in systemic lupus erythematosus. We also find TFs known for stress response, suggesting routine experimental caveats that warrant careful consideration. WhichTF yields biological insight into known and novel molecular mechanisms of TF-mediated transcriptional regulation in diverse contexts, including human and mouse cell types, cell fate trajectories, and disease-associated cells.

Acute appendicitis severity during the early COVID-19 pandemic period.

BACKGROUND: The early COVID-19 pandemic period significantly strained the US healthcare system. During this period, consultations and admissions for acute medical conditions decreased, which was associated with an increase in disease-specific morbidity and mortality. Therefore, we sought to determine what, if any, effect the early COVID-19 pandemic period had on the presentation, management, and histopathologic severity of acute appendicitis. METHODS: We performed a retrospective, observational study to compare the frequencies with which patients presented with acute appendicitis, the proportion of whom were managed surgically, and the distribution of histopathologic disease severity among all resected appendix specimens during the early COVID-19 pandemic period (March 6-June 30, 2020) to equivalent time periods for the 3 preceding/pre-pandemic years (2017-2019). RESULTS: Compared with equivalent pre-pandemic time periods, during the COVID-19 pandemic period there was no significant difference in the number of patients who presented for acute appendicitis, there was a decreased rate of surgical management (81% vs 94%; p=0.014), and there was an overall increase in the incidence of perforated appendicitis (31% vs 16%; p=0.004), including by histopathologic diagnosis (25% vs 11%; p=0.01). DISCUSSION: Despite potential patient hesitancy to present for care, the early COVID-19 pandemic period was associated with no significant change in the number of patients presenting with acute appendicitis; however, there was a significant increase in the incidence of perforated appendicitis. This study highlights the need to encourage patients to avoid late presentation for acute surgical conditions and for the robust planning for the medical management of otherwise surgical abnormalities during episodes of restricted or limited resources. LEVEL OF EVIDENCE: Level III.

Transcription factor expression defines subclasses of developing projection neurons highly similar to single-cell RNA-seq subtypes.

We are only just beginning to catalog the vast diversity of cell types in the cerebral cortex. Such categorization is a first step toward understanding how diversification relates to function. All cortical projection neurons arise from a uniform pool of progenitor cells that lines the ventricles of the forebrain. It is still unclear how these progenitor cells generate the more than 50 unique types of mature cortical projection neurons defined by their distinct gene-expression profiles. Moreover, exactly how and when neurons diversify their function during development is unknown. Here we relate gene expression and chromatin accessibility of two subclasses of projection neurons with divergent morphological and functional features as they develop in the mouse brain between embryonic day 13 and postnatal day 5 in order to identify transcriptional networks that diversify neuron cell fate. We compare these gene-expression profiles with published profiles of single cells isolated from similar populations and establish that layer-defined cell classes encompass cell subtypes and developmental trajectories identified using single-cell sequencing. Given the depth of our sequencing, we identify groups of transcription factors with particularly dense subclass-specific regulation and subclass-enriched transcription factor binding motifs. We also describe transcription factor-adjacent long noncoding RNAs that define each subclass and validate the function of Myt1l in balancing the ratio of the two subclasses in vitro. Our multidimensional approach supports an evolving model of progressive restriction of cell fate competence through inherited transcriptional identities.

Research Techniques Made Simple: Cutaneous Colorimetry: A Reliable Technique for Objective Skin Color Measurement.

Skin color evaluation contributes to assessment of an individual's cutaneous phenotype. Skin color changes provide important clues to disease progression or treatment response. Skin color is also a predictor of skin cancer risk. Melanin pigment, blood flow, skin thickness, and photoaging contribute to skin color. Melanin, hemoglobin, bilirubin, and carotene are the primary chromophores of skin color. Their concentrations vary depending on the individual's phenotype, anatomic location, external insults of chemical irritants and UVR, and physiological changes. The evaluation and perception of skin color are often subjective. Objective quantification of skin color can be achieved with colorimetric devices such as tristimulus colorimeters. These devices compute the intensity of light reflected from skin and correlate with pigmentation and erythema. Cutaneous color and color changes can be quantified under color organization systems, such as the CIELAB color space, which is standardized by the Commission Internationale de l'Eclairage (CIE). The CIELAB expresses color's lightness, red/green intensity, and yellow/blue intensity, as L*, a*, and b* values, respectively. Additionally, skin color's full spectral characteristics and cutaneous physiology can be measured with spectrophotometers. This article outlines basic principles of the CIELAB color system and how to optimally use colorimetric devices as a skin research tool.