Incidence and natural history of gastric high-grade dysplasia in patients with familial adenomatous polyposis syndrome.

BACKGROUND AND AIMS: Familial adenomatous polyposis (FAP) is characterized by high risks of colonic and extracolonic tumors. Recent studies have suggested a rising risk for gastric cancer (GC). We sought to define the spectrum of premalignant gastric polyps in FAP, focusing on high-grade dysplasia (HGD). METHODS: The gastric phenotypes of 118 patients diagnosed with FAP or attenuated FAP in our Hereditary Gastrointestinal Cancer Registry were retrospectively reviewed. To analyze the clinical features associated with the diagnosis of HGD, we established an age- and sex-matched control group of FAP patients from our cohort without gastric HGD in a 4:1 ratio. RESULTS: The spectrum and frequency of gastric polyps in individuals with FAP included fundic gland polyps (67.9%), hyperplastic polyps/foveolar hyperplasia (19.6%), tubular adenomas (15.2%), foveolar adenomas (10.7%), and pyloric gland adenomas (6.3%). Ten patients (8.9%) exhibited gastric HGD at a mean age of 55 ± 13 years, and HGD was seen in all polyp types. When compared with control subjects, HGD was associated with a high diversity of gastric polyp histology, prior low-grade dysplasia, severe gastric polyposis, and prior Whipple surgery (P = 2.0E-5, .003, .024, and .04, respectively). Two patients (1.7%) with HGD were diagnosed with GC. However, the remaining 8 patients with HGD have been under surveillance for an average of 5.8 ± 4.5 years without progression to GC. CONCLUSIONS: Gastric HGD in FAP may be more common than previously appreciated. The natural history of HGD is variable, and most patients with HGD do not appear to progress to GC.

Concordance of EUS and MRI/MRCP findings among high-risk individuals undergoing pancreatic cancer screening.

BACKGROUND/OBJECTIVES: Surveillance with endoscopic ultrasonography (EUS) and MRI/magnetic retrograde cholangiopancreatography (MRCP) is recommended for individuals at high risk for pancreatic cancer. We sought to characterize the findings of these surveillance exams and define the level of concordance between these two modalities. METHODS: 173 asymptomatic high-risk individuals (HRIs) meeting criteria for pancreatic cancer surveillance underwent EUS, MRI/MRCP, or both between 2008 and 2021. Clinical records were reviewed in all cases. RESULTS: HRIs underwent an average of 3.6 ± 3.2 surveillance exams over a period of 3.3 ± 3.5 years. Abnormalities including intraductal papillary mucinous neoplasms (IPMNs), solid lesions, and parenchymal irregularities were identified in 50.9% (n = 88). Four of these abnormalities (2.3%) had worrisome features, defined by cyst size, thickened/enhancing cyst walls, rapid growth rate, or change in main pancreatic duct diameter. All four worrisome lesions were seen on both MRI/MRCP and EUS. No pancreatic cancers were detected. Baseline EUS and MRI/MRCP exams were compared in 106 patients for concordance, and most (n = 66, 62.3%) were concordant. High levels of concordance were specifically observed for a dilated main pancreatic duct (p < 0.01) and cystic lesions >5 mm (p = 0.01). Among discordant cases, most (30/40; 75%) involved abnormal tissue heterogeneity seen primarily on EUS. None of these discordant lesions ultimately developed worrisome features. CONCLUSIONS: Worrisome pancreatic lesions were uncommon in our high-risk pancreatic cancer population and were detected by both EUS and MRI/MRCP. There was mild discordance with respect to less worrisome findings, but these discrepancies were not associated with any adverse clinical outcomes.

Coresets for the Average Case Error for Finite Query Sets.

Coreset is usually a small weighted subset of an input set of items, that provably approximates their loss function for a given set of queries (models, classifiers, hypothesis). That is, the maximum (worst-case) error over all queries is bounded. To obtain smaller coresets, we suggest a natural relaxation: coresets whose average error over the given set of queries is bounded. We provide both deterministic and randomized (generic) algorithms for computing such a coreset for any finite set of queries. Unlike most corresponding coresets for the worst-case error, the size of the coreset in this work is independent of both the input size and its Vapnik-Chervonenkis (VC) dimension. The main technique is to reduce the average-case coreset into the vector summarization problem, where the goal is to compute a weighted subset of the n input vectors which approximates their sum. We then suggest the first algorithm for computing this weighted subset in time that is linear in the input size, for n≫1/ε, where ε is the approximation error, improving, e.g., both [ICML'17] and applications for principal component analysis (PCA) [NIPS'16]. Experimental results show significant and consistent improvement also in practice. Open source code is provided.

No Fine-Tuning, No Cry: Robust SVD for Compressing Deep Networks.

A common technique for compressing a neural network is to compute the k-rank ℓ2 approximation Ak of the matrix A∈Rn×d via SVD that corresponds to a fully connected layer (or embedding layer). Here, d is the number of input neurons in the layer, n is the number in the next one, and Ak is stored in O((n+d)k) memory instead of O(nd). Then, a fine-tuning step is used to improve this initial compression. However, end users may not have the required computation resources, time, or budget to run this fine-tuning stage. Furthermore, the original training set may not be available. In this paper, we provide an algorithm for compressing neural networks using a similar initial compression time (to common techniques) but without the fine-tuning step. The main idea is replacing the k-rank ℓ2 approximation with ℓp, for p∈[1,2], which is known to be less sensitive to outliers but much harder to compute. Our main technical result is a practical and provable approximation algorithm to compute it for any p≥1, based on modern techniques in computational geometry. Extensive experimental results on the GLUE benchmark for compressing the networks BERT, DistilBERT, XLNet, and RoBERTa confirm this theoretical advantage.

Autonomous Toy Drone via Coresets for Pose Estimation.

A coreset of a dataset is a small weighted set, such that querying the coreset provably yields a ( 1 + ε )-factor approximation to the original (full) dataset, for a given family of queries. This paper suggests accurate coresets ( ε = 0 ) that are subsets of the input for fundamental optimization problems. These coresets enabled us to implement a "Guardian Angel" system that computes pose-estimation in a rate > 20 frames per second. It tracks a toy quadcopter which guides guests in a supermarket, hospital, mall, airport, and so on. We prove that any set of n matrices in R d × d whose sum is a matrix S of rank r, has a coreset whose sum has the same left and right singular vectors as S, and consists of O ( d r ) = O ( d 2 ) matrices, independent of n. This implies the first (exact, weighted subset) coreset of O ( d 2 ) points to problems such as linear regression, PCA/SVD, and Wahba's problem, with corresponding streaming, dynamic, and distributed versions. Our main tool is a novel usage of the Caratheodory Theorem for coresets, an algorithm that computes its set in time that is linear in its cardinality. Extensive experimental results on both synthetic and real data, companion video of our system, and open code are provided.

Magnesium Deficiency and Minimal Hepatic Encephalopathy among Patients with Compensated Liver Cirrhosis.

BACKGROUND: Magnesium is an essential intracellular cation. Magnesium deficiency is common in the general population and its prevalence among patients with cirrhosis is even higher. Correlation between serum levels and total body content is poor because most magnesium is intracellular. Minimal hepatic encephalopathy is a subclinical phase of hepatic encephalopathy with no overt symptoms. Cognitive exams can reveal minor changes in coordination, attention, and visuomotor function, whereas language and verbal intelligence are usually relatively spared. OBJECTIVES: To assess the correlation between intracellular and serum magnesium levels and minimal hepatic encephalopathy. METHODS: Outpatients with a diagnosis of compensated liver cirrhosis were enrolled in this randomized, double-blinded study. Patients were recruited for the study from November 2013 to January 2014, and were randomly assigned to a control (placebo) or an interventional (treated with magnesium oxide) group. Serum and intracellular magnesium levels were measured at enrollment and at the end of the study. Cognitive function was assessed by a specialized occupational therapist. RESULTS: Forty-two patients met the inclusion criteria, 29 of whom were included in this study. Among these, 83% had abnormal cognitive exam results compatible with minimal hepatic encephalopathy. While only 10% had hypomagnesemia, 33.3% had low levels of intracellular magnesium. Initial intracellular and serum magnesium levels positively correlated with cognitive performance. CONCLUSIONS: Magnesium deficiency is common among patients with compensated liver cirrhosis. We found an association between magnesium deficiency and impairment in several cognitive function tests. This finding suggests involvement of magnesium in the pathophysiology of minimal hepatic encephalopathy.

Treatment of esophagogastric junction outflow obstruction with concomitant hypercontractile esophagus: A case series.

Hypercontractile esophagus with concomitant esophagogastric junction outflow obstruction (EGJOO) is a rare entity that is characterized by both esophageal hypercontractility and lack of relaxation of the EGJ. The clinical characteristics of these patients are not well-described and there is no strict recommendation regarding the treatment of this condition. We report four cases of patients with hypercontractile esophagus and concomitant to EGJOO. All patients underwent upper gastrointestinal (GI) endoscopy, high-resolution esophageal manometry (HRM) and barium swallow and met the criteria of Chicago Classification for both EGJOO and hypercontractile esophagus. Patients were followed up to four years from diagnosis and clinical symptoms were recorded. Four patients, who underwent evaluation for dysphagia, were found to have both EGJOO and hypercontractile esophagus on HRM. Two of them had mild symptoms and did not undergo treatment with no progression of symptoms on follow-up. Of the two patients who underwent treatment, one had botulinum toxin injection to the EGJ via upper GI endoscopy and one underwent per-oral endoscopic myotomy. Symptoms in both patients improved. Patients with concomitant hypercontractile esophagus and EGJOO present with varying degrees of symptoms and the treatment approach should be personalized according to the degree of symptoms and general clinical condition.

Fast and Accurate Least-Mean-Squares Solvers for High Dimensional Data.

Least-mean-squares (LMS) solvers such as Linear / Ridge-Regression and SVD not only solve fundamental machine learning problems, but are also the building blocks in a variety of other methods, such as matrix factorizations. We suggest an algorithm that gets a finite set of n d-dimensional real vectors and returns a subset of d+1 vectors with positive weights whose weighted sum is exactly the same. The constructive proof in Caratheodory's Theorem computes such a subset in O(n2d2) time and thus not used in practice. Our algorithm computes this subset in O(nd+d4logn) time, using O(logn) calls to Caratheodory's construction on small but "smart" subsets. This is based on a novel paradigm of fusion between different data summarization techniques, known as sketches and coresets. For large values of d, we suggest a faster construction that takes O(nd) time and returns a weighted subset of O(d) sparsified input points. Here, a sparsified point means that some of its entries were set to zero. As an application, we show how to boost the performance of existing LMS solvers, such as those in scikit-learn library, up to x100. Generalization for streaming and distributed data is trivial. Extensive experimental results and open source code are provided.

A Unified Approach to Coreset Learning.

Coreset of a given dataset and loss function is usually a small weighed set that approximates this loss for every query from a given set of queries. Coresets have shown to be very useful in many applications. However, coresets' construction is done in a problem-dependent manner and it could take years to design and prove the correctness of a coreset for a specific family of queries. This could limit coresets' use in practical applications. Moreover, small coresets provably do not exist for many problems. To address these limitations, we propose a generic, learning-based algorithm for construction of coresets. Our approach offers a new definition of coreset, which is a natural relaxation of the standard definition and aims at approximating the average loss of the original data over the queries. This allows us to use a learning paradigm to compute a small coreset of a given set of inputs with respect to a given loss function using a training set of queries. We derive formal guarantees for the proposed approach. Experimental evaluation on deep networks and classic machine learning problems show that our learned coresets yield comparable or even better results than the existing algorithms with worst case theoretical guarantees (that may be too pessimistic in practice). Furthermore, our approach applied to deep network pruning provides the first coreset for a full deep network, i.e., compresses all the networks at once, and not layer by layer or similar divide-and-conquer methods.

Data-Independent Structured Pruning of Neural Networks via Coresets.

Model compression is crucial for the deployment of neural networks on devices with limited computational and memory resources. Many different methods show comparable accuracy of the compressed model and similar compression rates. However, the majority of the compression methods are based on heuristics and offer no worst case guarantees on the tradeoff between the compression rate and the approximation error for an arbitrarily new sample. We propose the first efficient structured pruning algorithm with a provable tradeoff between its compression rate and the approximation error for any future test sample. Our method is based on the coreset framework, and it approximates the output of a layer of neurons/filters by a coreset of neurons/filters in the previous layer and discards the rest. We apply this framework in a layer-by-layer fashion from the bottom to the top. Unlike previous works, our coreset is data-independent, meaning that it provably guarantees the accuracy of the function for any input [Formula: see text], including an adversarial one.

Clinical features and long-term outcomes of patients with colonic oligopolyposis of unknown etiology.

BACKGROUND: Colonic adenomatous polyposis of unknown etiology (CPUE) is an adenomatous polyposis phenotype that resembles Familial Adenomatous Polyposis (FAP) even though no germline pathogenic variant is identified. AIM: We sought to better characterize the clinical features and outcomes in a cohort of CPUE patients. METHODS: This is a retrospective case series of patients 18 years old or older with aden-omatous oligopolyposis (between 10-100 adenomas) and negative genetic testing, identified through the Hereditary Gastrointestinal Cancer Database at Massachusetts General Hospital, a tertiary academic referral center. A retrospective chart review was performed with a focus on demographics, alcohol and tobacco use, medication use, familial malignancy and polyp burden, genetic testing information, endoscopic surveillance data including the corresponding histopathology, colonic and extracolonic malignancies, mortality events, and their etiology. Spearman correlation and Pearson Chi-square test (or Fisher's exact test) were used for continuous and categorical variables respectively. RESULTS: CPUE patients were primarily male (69%) and presented for genetic counseling at 63.7 years. Only 2 patients (2.9%) reported a first-degree relative with polyposis. During an average surveillance period of 12.3 years, 0.5 colonoscopies per year were performed. Patients developed 2.3 new adenomas per year. 4 (5.7%) were diagnosed with colorectal cancer (CRC) at a mean age of 66 years, and 3 were diagnosed prior to the onset of oligopolyposis. 7 (10%) required colectomy due to advanced dysplasia or polyp burden. With respect to upper gastrointestinal manifestations, 1 patient had a gastric adenoma, but there were no cases of gastric or small bowel polyposis. During surveillance, 10 (14%) patients died at a mean age of 72, and none were due to CRC. CONCLUSION: CPUE is distinct from familial adenomatous polyposis (FAP) syndrome and the use of FAP surveillance guidelines may result in unnecessarily frequent upper and lower endoscopies.

Characterization of blood-derived exosomal hTERT mRNA as a biomarker for colon cancer and Lynch syndrome.

Background: Human telomerase reverse transcriptase (hTERT)- mRNA was shown to be elevated in exosomes derived from the sera of a variety of hematological and solid cancer patients. We aimed to evaluate its role as a diagnostic marker in patients with newly diagnosed colon cancer and in hereditary syndromes with predisposition to colon cancer. Methods: hTERT -mRNA levels were determined in serum-derived exosomes from 88 patients with colon cancer, 71 Lynch-syndrome carriers with unknown active malignancies and 50 healthy controls. Data, including demographics, background diseases, clinical data regarding tumor characteristics and genetic data, were retrieved data from medical files. Results: Patients with colon cancer had both higher exosomal hTERT mRNA levels and a higher proportion of patients with positive exosomal hTERT mRNA than controls (29.5% vs. 4%, respectively, P values < 0.001). Within the cancer group, patients with a metastatic disease had higher levels of telomerase mRNA than non-metastatic disease patients, and these levels correlated with CEA levels. Likewise, Lynch syndrome carriers had a higher proportion of positive exosomal hTERT mRNA than controls (21.1% vs. 4%, respectively, P value 0.008) but only a trend towards higher exosomal hTERT mRNA levels. Higher telomerase mRNA levels were not correlated with the mutated gene. Conclusions: Exosomal serum hTERT -mRNA levels are associated with metastatic colon cancer and were also demonstrated in a subset of Lynch syndrome carriers. Its significance as a biomarker for developing malignancy should be elucidated.

Genetic testing for assessment of lynch syndrome in young patients with polyps.

BACKGROUND: Routine screening for establishing Lynch syndrome (LS) in young individuals diagnosed with adenomas is not recommended due to its low yield, and limited sensitivity of the employment of immunohistochemistry for DNA mismatch-repair proteins on polyps. Hence we aimed to evaluate the yield of germline mutational analysis in diagnosis of LS in a young Israeli cohort with colorectal adenomatous polyps. METHODS: Data were retrospectively collected on consecutive patients, age ≤ 45 years, who underwent colonoscopy with removal of at least one adenoma during 2015-2020, and subsequently genetic testing by multigene panel or LS-Jewish founder mutation panel. RESULTS: Overall, 92 patients were included (median age 35 years, range 23-45 years), of whom 79 (85.8%) underwent multigene panel genotyping, and 13 (14.2%) analysis for Jewish founder LS gene mutations. Altogether, 18 patients were identified with pathogenic mutations in actionable genes, including LS-associated genes in 6 (6.5%), BRCA2 in 2 (2.5%), GREM1 in 1(1.2%), and low-penetrance genes- APC I1307K and CHEK2- in 9 (11.4%) patients. Compared with non-LS patients, LS-carriers had a significantly higher median PREMM5 score (2.6 vs. 1.3; P = 0.04). CONCLUSIONS: Young individuals diagnosed with adenomatous polyps should be offered genetic testing when fulfilling clinical guidelines for LS, but weight should also be given to adenoma characteristics in the PREMM5 score.

Computerized Tomography Criteria as a Tool for Simplifying the Assessment of Locally Advanced Rectal Cancer.

BACKGROUND: Rectal cancer represents a leading cause of mortality worldwide. Staging defines the local and distant extent of the disease, guides management, and predicts prognosis. Different modalities are available for staging including TRUS (transrectal ultrasound), CT (computed tomography), and MRI (magnetic resonance imaging). OBJECTIVE: The objective of this study was to screen and isolate CT imaging parameters suggestive of advanced rectal cancer and its utility as a tool in simplifying the staging protocol making further imaging studies unnecessary. DESIGN: Retrospective, single center study. PATIENTS AND SETTINGS: Seventy-five patients with rectal carcinoma were included and were divided into two groups according to their T score and nodal involvement status, as diagnosed by TRUS. Group 1 (n = 15) "local disease" (T1/T2 N0) and group 2 (n = 60) "locally advanced disease" are both eligible for neoadjuvant treatment (N/any T or T3/any N). For each patient, three CT imaging parameters that represent locally advanced disease, i.e., perirectal fat infiltration, local lymphadenopathy, and rectal wall thickening, were evaluated and compared between the two groups. MAIN OUTCOME MEASURE: The capability of CT imaging to accurately predict locally advanced rectal carcinoma. RESULTS: Rectal wall thickening on CT was found to have 92% PPV and perirectal lymphadenopathy 96% PPV for predicting a locally advanced stage. A combination of those two parameters results in a predictive PPV of 98%. LIMITATIONS: This was a single center retrospective study, with a relatively small cohort. CONCLUSIONS: CT is a valuable tool in the assessment and management of rectal carcinoma as it can identify locally advanced rectal cancer. This enables treatment without any further unnecessary evaluation.

Estimation of Gastric Volume Before Anesthesia in Term-Pregnant Women Undergoing Elective Cesarean Section, Compared With Non-pregnant or First-Trimester Women Undergoing Minor Gynecological Surgical Procedures.

BACKGROUND: Traditionally, intubation of pregnant women has been performed using a rapid sequence induction. This is due to the classical concept that women with more than 18 weeks of pregnancy (mid-second trimester) are always considered to have an increased risk of aspiration due to a number of factors, regardless of the fasting duration. Rapid sequence induction is associated with a higher rate of adverse events. AIMS: Our study aimed to illuminate the hypothesis that there is no difference in gastric volume between term-pregnant women and non-pregnant or first-trimester pregnant women who were undergoing minor gynecological surgical procedures. Accordingly, we measured gastric volume and content before anesthesia in term-pregnant women undergoing elective cesarean section, and to compare it with non-pregnant or first-trimester pregnant women who were undergoing minor gynecological surgical procedures. METHODS: In this single-center prospective study, the gastric volume and content were assessed by abdominal ultrasound (AUS) just prior to the scheduled procedure. AUS was performed in the sagittal or para-sagittal plain in the upright position and the stomach content was estimated according to the antral circumferential area. Group 1 consisted of 50 term-pregnant women scheduled for cesarean section. Group 2 consisted of 45 non-pregnant or first-trimester pregnant women who were scheduled for minor gynecologic procedure. RESULTS: Despite significant longer fasting time prior to the interventional procedure in the non-pregnant or first-trimester women group, there was no significant difference in gastric volume between term-pregnant and first-trimester pregnant women (3.2 ± 0.97 cm2 vs 3.2 ± 0.79 cm2; P = .97). Gastric volume was small in the two groups. CONCLUSION: Fasting gastric volume before cesarean section in term-pregnant women is small and is not different than in non-pregnant or first-trimester women undergoing minor gynecologic procedures. Ultrasound estimation of gastric volume is a reliable and easy-to-perform technique which might help in decision-making regarding the airway management prior to induction of anesthesia in pregnant women.

Ocular dominance plasticity in mature mice.

Ocular dominance plasticity, classically thought to be restricted to an early critical period, is now described by Sawtell et al. in fully adult mice. Adult plasticity, like critical period plasticity, requires cortical NMDA receptors but involves different functional changes in cortical circuits.

Spike timing-dependent synaptic depression in the in vivo barrel cortex of the rat.

Spike timing-dependent plasticity (STDP) is a computationally powerful form of plasticity in which synapses are strengthened or weakened according to the temporal order and precise millisecond-scale delay between presynaptic and postsynaptic spiking activity. STDP is readily observed in vitro, but evidence for STDP in vivo is scarce. Here, we studied spike timing-dependent synaptic depression in single putative pyramidal neurons of the rat primary somatosensory cortex (S1) in vivo, using two techniques. First, we recorded extracellularly from layer 2/3 (L2/3) and L5 neurons, and paired spontaneous action potentials (postsynaptic spikes) with subsequent subthreshold deflection of one whisker (to drive presynaptic afferents to the recorded neuron) to produce "post-leading-pre" spike pairings at known delays. Short delay pairings (<17 ms) resulted in a significant decrease of the extracellular spiking response specific to the paired whisker, consistent with spike timing-dependent synaptic depression. Second, in whole-cell recordings from neurons in L2/3, we paired postsynaptic spikes elicited by direct-current injection with subthreshold whisker deflection to drive presynaptic afferents to the recorded neuron at precise temporal delays. Post-leading-pre pairing (<33 ms delay) decreased the slope and amplitude of the PSP evoked by the paired whisker, whereas "pre-leading-post" delays failed to produce depression, and sometimes produced potentiation of whisker-evoked PSPs. These results demonstrate that spike timing-dependent synaptic depression occurs in S1 in vivo, and is therefore a plausible plasticity mechanism in the sensory cortex.

Melanosis Coli: A Helpful Contrast Effect or a Harmful Pigmentation?

BACKGROUND: Melanosis coli, a brown discoloration of colonic mucosa, is considered as a benign condition mainly observed in patients under chronic anthranoid laxatives. Recent data link this condition with an increased adenoma detection rate. Moreover, its tumorigenic potential and possible association with the development of colorectal cancer remains uncertain. We conducted this study to compare the polyp detection rate and colorectal cancer diagnosis in patients with melanosis against matched control group without melanosis. PATIENTS AND METHODS: A retrospective single-center study. Patients diagnosed with melanosis coli on colonoscopy over a 15-year period were included. Each melanosis coli patient was matched with three controls by age, gender, setting (inpatient/outpatient), and procedure's indication. Polyp detection rate and diagnosis of colorectal cancer were recorded and compared between the groups before and after adjustment for bowel preparation. RESULTS: A cohort of 718 patients with melanosis and 2154 controls were included. The polyp detection rates were 33.4% and 21.8% of melanosis and control groups, respectively (P < .001). Melanosis coli, however, was associated with less diagnosis of colorectal cancer than controls (0.3% vs 3.9%; P < .001). In multivariate analysis, melanosis diagnosis on endoscopy was significantly associated with higher polyp detection rate (odds ratio [OR] = 1.986, 95% confidence interval [CI]: 1.626-2.425; P value < .01). CONCLUSIONS: Melanosis coli is not associated with increased diagnosis of colorectal cancer. It is associated, however, with enhanced polyp detection likely due to chromo-endoscopy-like effect.