A Convolutional-Transformer Model for FFR and iFR Assessment From Coronary Angiography.

The quantification of stenosis severity from X-ray catheter angiography is a challenging task. Indeed, this requires to fully understand the lesion's geometry by analyzing dynamics of the contrast material, only relying on visual observation by clinicians. To support decision making for cardiac intervention, we propose a hybrid CNN-Transformer model for the assessment of angiography-based non-invasive fractional flow-reserve (FFR) and instantaneous wave-free ratio (iFR) of intermediate coronary stenosis. Our approach predicts whether a coronary artery stenosis is hemodynamically significant and provides direct FFR and iFR estimates. This is achieved through a combination of regression and classification branches that forces the model to focus on the cut-off region of FFR (around 0.8 FFR value), which is highly critical for decision-making. We also propose a spatio-temporal factorization mechanisms that redesigns the transformer's self-attention mechanism to capture both local spatial and temporal interactions between vessel geometry, blood flow dynamics, and lesion morphology. The proposed method achieves state-of-the-art performance on a dataset of 778 exams from 389 patients. Unlike existing methods, our approach employs a single angiography view and does not require knowledge of the key frame; supervision at training time is provided by a classification loss (based on a threshold of the FFR/iFR values) and a regression loss for direct estimation. Finally, the analysis of model interpretability and calibration shows that, in spite of the complexity of angiographic imaging data, our method can robustly identify the location of the stenosis and correlate prediction uncertainty to the provided output scores.

Neural Transformers for Intraductal Papillary Mucosal Neoplasms (IPMN) Classification in MRI images.

Early detection of precancerous cysts or neoplasms, i.e., Intraductal Papillary Mucosal Neoplasms (IPMN), in pancreas is a challenging and complex task, and it may lead to a more favourable outcome. Once detected, grading IPMNs accurately is also necessary, since low-risk IPMNs can be under surveillance program, while high-risk IPMNs have to be surgically resected before they turn into cancer. Current standards (Fukuoka and others) for IPMN classification show significant intra- and inter-operator variability, beside being error-prone, making a proper diagnosis unreliable. The established progress in artificial intelligence, through the deep learning paradigm, may provide a key tool for an effective support to medical decision for pancreatic cancer. In this work, we follow this trend, by proposing a novel AI-based IPMN classifier that leverages the recent success of transformer networks in generalizing across a wide variety of tasks, including vision ones. We specifically show that our transformer-based model exploits pre-training better than standard convolutional neural networks, thus supporting the sought architectural universalism of transformers in vision, including the medical image domain and it allows for a better interpretation of the obtained results.

TwinLiverNet: Predicting TACE Treatment Outcome from CT scans for Hepatocellular Carcinoma using Deep Capsule Networks.

Predicting response to treatment plays a key role to assist radiologists in hepato-cellular carcinoma (HCC) therapy planning. The most widely used treatment for unresectable HCC is the trans-arterial chemoembolization (TACE). A complete radiological response after the first TACE is a reliable predictor of treatment favourable outcome. However, visual inspection of contrast-enhanced CT scans is time-consuming, error prone and too operator-dependent. Thus, in this paper we propose TwinLiverNet: a deep neural network that is able to predict TACE treatment outcome through learning visual cue from CT scans. TwinLiverNet, specifically, integrates 3D convolutions and capsule networks and is designed to process simultaneously late arterial and delayed phases from contrast-enhanced CTs. Experimental results carried out on a dataset consisting of 126 HCC lesions show that TwinLiverNet reaches an average accuracy of 82% in predicting complete response to TACE treatment. Furthermore, combining multiple CT phases (specifically, late arterial and delayed ones) yields a performance increase of over 12 percent points. Finally, the introduction of capsule layers into the model avoids the model to overfit, while enhancing accuracy.Clinical relevance- TwinLiverNet supports radiologists in visual inspection of CT scans to assess TACE treatment outcome, while reducing inter-operator variability.

Deep learning for automated skeletal bone age assessment in X-ray images.

Skeletal bone age assessment is a common clinical practice to investigate endocrinology, genetic and growth disorders in children. It is generally performed by radiological examination of the left hand by using either the Greulich and Pyle (G&P) method or the Tanner-Whitehouse (TW) one. However, both clinical procedures show several limitations, from the examination effort of radiologists to (most importantly) significant intra- and inter-operator variability. To address these problems, several automated approaches (especially relying on the TW method) have been proposed; nevertheless, none of them has been proved able to generalize to different races, age ranges and genders. In this paper, we propose and test several deep learning approaches to assess skeletal bone age automatically; the results showed an average discrepancy between manual and automatic evaluation of about 0.8 years, which is state-of-the-art performance. Furthermore, this is the first automated skeletal bone age assessment work tested on a public dataset and for all age ranges, races and genders, for which the source code is available, thus representing an exhaustive baseline for future research in the field. Beside the specific application scenario, this paper aims at providing answers to more general questions about deep learning on medical images: from the comparison between deep-learned features and manually-crafted ones, to the usage of deep-learning methods trained on general imagery for medical problems, to how to train a CNN with few images.

Factors affecting the oligomeric state of NADP-malic enzyme from maize and wheat tissues: a chemical crosslinking study.

The different aggregational states of maize and wheat NADP-malic enzyme as affected by pH, temperature and various metabolites have been studied by the combined use of intersubunit crosslinking and denaturing polyacrylamide gel electrophoresis. The association/dissociation equilibrium is a pH-dependent process: pH values above 8.0 promote the tetramer formation, while lowering the pH shifts the equilibria towards dimers and monomers. Below pH 6.0, most molecules exist as monomers. In the same way, the temperature governs the equilibria between the different oligomeric states. As the temperature is lowered from 42 to 0 degrees C, a progressive dissociation into dimers and monomers is observed. Excess enthalpies are negative in all cases, but the overall process demands an input of Gibb's free energy. Consequently, the protein dissociation is an entropy-driven process. The presence of Mg2+ or glycerol induces aggregation in both enzymes, while increasing the ionic strength produces the opposite effect. The results suggest that changes in the equilibria between monomer, dimer and tetramer of NADP-malic enzyme could be the molecular basis for an effective regulation of the enzyme activity in vivo.

Transcranial magnetic stimulation in the assessment of motor cortex excitability and treatment of drug-resistant major depression.

Major depression is one of the leading causes of disabling condition worldwide and its treatment is often challenging and unsatisfactory, since many patients become refractory to pharmacological therapies. Transcranial magnetic stimulation (TMS) is a noninvasive neurophysiological investigation mainly used to study the integrity of the primary motor cortex excitability and of the cortico-spinal tract. The development of paired-pulse and repetitive TMS (rTMS) paradigms has allowed investigators to explore the pathophysiology of depressive disorders and other neuropsychiatric diseases linked to brain excitability dysfunctions. Repetitive transcranial magnetic stimulation has also therapeutic and rehabilitative capabilities since it is able to induce changes in the excitability of inhibitory and excitatory neuronal networks that may persist in time. However, the therapeutic effects of rTMS on major depression have been demonstrated by analyzing only the improvement of neuropsychological performance. The aim of this study was to investigate cortical excitability changes on 12 chronically-medicated depressed patients (test group) after rTMS treatment and to correlate neurophysiological findings to neuropsychological outcomes. In detail, we assessed different parameters of cortical excitability before and after active rTMS in the test group, then compared to those of 10 age-matched depressed patients (control group) who underwent sham rTMS. In line with previous studies, at baseline both groups exhibited a significant interhemispheric difference of motor cortex excitability. This neurophysiological imbalance was then reduced in the patients treated with active rTMS, resulting also in a clinical benefit as demonstrated by the improvement in neuropsychological test scores. On the contrary, after sham rTMS, the interhemispheric difference was still evident in the control group. The reported clinical benefits in the test group might be related to the plastic remodeling of synaptic connection induced by rTMS treatment.

An integrated computer-controlled system for assisting researchers in cortical excitability studies by using transcranial magnetic stimulation.

Transcranial magnetic stimulation (TMS) is the most important technique currently available to study cortical excitability. Additionally, TMS can be used for therapeutic and rehabilitation purposes, replacing the more painful transcranial electric stimulation (TES). In this paper we present an innovative and easy-to-use tool that enables neuroscientists to design, carry out and analyze scientific studies based on TMS experiments for both diagnostic and research purposes, assisting them not only in the practicalities of administering the TMS but also in each step of the entire study's workflow. One important aspect of this tool is that it allows neuroscientists to specify research designs at will, enabling them to define any parameter of a TMS study starting from data acquisition and sample group definition to automated statistical data analysis and RDF data storage. It also supports the diagnosing process by using on-line support vector machines able to learn incrementally from the diseases instances that are continuously added into the system. The proposed system is a neuroscientist-centred tool where the protocols being followed in TMS studies are made explicit, leaving to the users flexibility in exploring and sharing the results, and providing assistance in managing the complexity of the final diagnosis. This type of tool can make the results of medical experiments more easily exploitable, thus accelerating scientific progress.

Epiphysis and metaphysis extraction and classification by adaptive thresholding and DoG filtering for automated skeletal bone age analysis.

The aim of this work is to develop an automated system for skeletal bone age evaluation. Skeletal bone age assessment is a common and time-consuming task in pediatric radiology. We describe a system that implements the EMROI (Epiphyseal/Metaphyseal Region Of Interest) extraction method by image processing techniques. Each EMROI is localized on the image, by using an analysis of its gray levels. Then the bones in the EMROI are extracted by using the DoG (Difference of Gaussians) filter and enhanced using a novel adaptive thresholding obtained by histogram processing. Finally, the main features of these bones are extracted for the stage TW2 (Tanner Whitehouse) evaluation. Results obtained on a sample of 20 X-rays are discussed. With respect to the other existing automated approaches to compute the bone age, the proposed method requires less user feedback on system's settings and is very versatile since it doesn't depend strongly on features of the X-ray acquisition.

An automated tool for face recognition using visual attention and active shape models analysis.

An entirely automated approach for the recognition of the face of a people starting from her/his images is presented. The approach uses a computational attention module to find automatically the most relevant facial features using the Focus Of Attentions (FOA) These features are used to build the model of a face during the learning phase and for recognition during the testing phase. The landmarking of the features is performed by applying the active contour model (ACM) technique, whereas the active shape model (ASM) is adopted for constructing a flexible model of the selected facial features. The advantages of this approach and opportunities for further improvements are discussed.

A multi-facets analysis of the driver status by EEG and fuzzy hardware processing.

In the paper, EEGs are used to perform a multi facets analysis of the driver status. The EEG tracks, taken by means of electrodes installed in a basket dressed by the driver, are processed by a fuzzy model consisting of rules able to predict possible temporary driver attention deficit due to stress or disease conditions. The driving behavior is evaluated in real time by a hardware fuzzy processing. The possibility of taking into account different facets of the driver status is claimed to give rise to a driver control system with good safety and predictive features.

The MutL ATPase is required for mismatch repair.

Members of the MutL family contain a novel nucleotide binding motif near their amino terminus, and the Escherichia coli protein has been found to be a weak ATPase (Ban, C., and Yang, W. (1998) Cell 95, 541-552). Genetic analysis has indicated that substitution of Lys for Glu-32 within this motif of bacterial MutL results in a strong dominant negative phenotype (Aronshtam, A., and Marinus, M. G. (1996) Nucleic Acids Res. 24, 2498-2504). By in vitro comparison of MutL-E32K with the wild type protein, we show the mutant protein to be defective in DNA-activated ATP hydrolysis, as well as MutS- and MutL-dependent activation of the MutH d(GATC) endonuclease and the mismatch repair excision system. MutL-E32K also acts in dominant negative manner in the presence of wild type MutL in vitro, inhibiting the overall mismatch repair reaction, as well as MutH activation. As judged by protein affinity chromatography, MutL and MutL-E32K both support formation of ternary complexes that also contain MutS and MutH or MutS and DNA helicase II. These findings imply that the MutL nucleotide binding center is required for mismatch repair and suggest that the dominant negative behavior of the MutL-E32K mutation is due to the formation of dead-end complexes in which the MutL-E32K protein is unable to transduce a signal from MutS that otherwise results in mismatch-dependent activation of the MutH d(GATC) endonuclease or the unwinding activity of helicase II.

Kinetic mechanism of NADP-malic enzyme from maize leaves.

The kinetic mechanism of NADP-dependent malic enzyme purified from maize leaves was studied in the physiological direction. Product inhibition and substrate analogues studies with 3' aminopyridine dinucleotide phosphate and tartrate indicate that the enzyme reaction follows a sequential ordered Bi-Ter kinetic mechanism. NADP is the leading substrate followed by L-malate and the products are released in the order of CO2, pyruvate and NADPH. The enzyme also catalyzes a slow, magnesium-dependent decarboxylation of oxaloacetate and reduction of pyruvate and oxaloacetate in the presence of NADPH to produce L-lactate and L-malate, respectively.

NADP(+)-malic enzyme from sugarcane leaves: structural properties studied by thermal inactivation.

The irreversible thermal inactivation of the sugarcane leaf NADP(+)-malic enzyme was studied at 50 degrees C and pH 7.0 and 8.0. Depending on the preincubation conditions, thermal inactivation followed mono- or biphasic first-order kinetics. A two-step behavior in the irreversible denaturation process was found when protein concentration was sufficiently low. The protein concentration necessary to obtain monlphasic thermal inactivation kinetics was lower at pH 8.0 than at pH 7.0. The results suggest that biphasic inactivation kinetics are the consequence of the existence of two different oligomeric forms of the enzyme (dimer and tetramer), with the dimer being more stable in regards to thermal inactivation. The effects of the substrate and essential cofactors on the thermostability and equilibrium between the dimeric and tetrameric enzyme forms were also studied. Depending on the pH, NADP+, L-malate, and Mg2+ all had a protective effect on the stability of the dimeric and tetrameric species during thermal treatment. However, these ligands showed different effects on the aggregation state of the enzyme. NADP+ and L-malate induced dissociation, especially at pH 8.0, whereas Mg2+ induced aggregation of the protein. By studying the thermal inactivation kinetics at 50 degrees C and different pH values it was observed that the equilibrium between dimers and tetramers was dramatically affected in the range of pH 7.0-8.0. These results suggest that an amino acid residue(s) in the protein with an apparent pKa value of 7.7 needs to be deprotonated to stabilize aggregation of the enzyme to the tetrameric form.

Interaction of analogues of substrate with NADP-malic enzyme from maize leaves.

The effect of structural analogues of L-malate was studied on NADP-malic enzyme purified from Zea mays L. leaves. Among the compounds tested, the organic acids behaved as more potent inhibitors at pH 7.0 than at pH 8.0, suggesting that the dimeric form was more susceptible to the inhibition than the tetrameric form of the enzyme.Oxalate, ketomalonate, hydroxymalonate, malonate, oxaloacetate, tartrate, α-hydroxybutyrate, α-ketobutyrate, α-ketoglutarate and α-hydroxyglutarate exhibited linear competitive inhibition with respect to the substrate L-malate at pH 8.0. On the other hand, glyoxylate and glycolate turned out to be non-competitive inhibitors, while glycolaldehyde, succinate, fumarate, maleate and ß- and γ-hydroxybutyrate had no effect on the enzyme activity, at the concentrations assayed. These results suggest that the extent of inhibition was dependent on the size of the analogues and that the presence of an 1-carboxyl group along with a 2-hydroxyl or 2-keto group was important for binding of the substrate analogue to the enzyme.

Evidence for the Existence of Two Essential and Proximal Cysteinyl Residues in NADP-Malic Enzyme from Maize Leaves.

Incubation of maize (Zea mays) leaf NADP-malic enzyme with monofunctional and bifunctional N-substituted maleimides results in an irreversible inactivation of the enzyme. Inactivation by the monofunctional reagents, N-ethylmaleimide (NEM) and N-phenylmaleimide, followed pseudo-first-order kinetics. The maximum inactivation rate constant for phenylmaleimide was 10-fold higher than that for NEM, suggesting a possible hydrophobic microenvironment of the residue(s) involved in the modification of the enzyme. In contrast, the inactivation kinetics with the bifunctional maleimides, ortho-, meta-, and para-phenylenebismaleimide, were biphasic, probably due to different reactivities of the groups reacting with the two heads of these bifunctional reagents, with a possible cross-linking of two sulfhydryl groups. The inactivation by mono and bifunctional maleimides was partially prevented by Mg(2+) and l-malate, and NADP prevented the inactivation almost totally. Determination of the number of reactive sulfhydryl groups of the native enzyme with [(3)H]NEM in the absence or presence of NADP showed that inactivation occurred concomitantly with the modification of two cysteinyl residues per enzyme monomer. The presence of these two essential residues was confirmed by titration of sulfhydryl groups with [(3)H]NEM in the enzyme previously modified by o-phenylenebismaleimide in the absence or presence of NADP.

Analogues of NADP(+) as inhibitors and coenzymes for NADP(+) malic enzyme from maize leaves.

Structural analogues of the NADP(+) were studied as potential coenzymes and inhibitors for NADP(+) dependent malic enzyme from Zea mays L. leaves. Results showed that 1, N(6)-etheno-nicotinamide adenine dinucleotide phosphate (∈ NADP(+)), 3-acetylpyridine-adenine dinucleotide phosphate (APADP(+)), nicotinamide-hypoxanthine dinucleotide phosphate (NHDP(+)) and ß-nicotinamide adenine dinucleotide 2': 3'-cyclic monophosphate (2'3'NADPc(+)) act as alternate coenzymes for the enzyme and that there is little variation in the values of the Michaelis constants and only a threefold variation in Vmax for the five nucleotides. On the other hand, thionicotinamide-adenine dinucleotide phosphate (SNADP(+)), 3-aminopyridine-adenine dinucleotide phosphate (AADP(+)), adenosine 2'-monophosphate (2'AMP) and adenosine 2': 3'-cyclic monophosphate (2'3'AMPc) were competitive inhibitors with respect to NADP(+), while ß-nicotinamide adenine dinucleotide 3'-phosphate (3'NADP(+)), NAD(+), adenosine 3'-monophosphate (3'AMP), adenosine 2': 5'-cyclic monophosphate (2'5'AMPc), 5'AMP, 5'ADP, 5'ATP and adenosine act as non-competitive inhibitors. These results, together with results of semiempirical self-consistent field-molecular orbitals calculations, suggest that the 2'-phosphate group is crucial for the nucleotide binding to the enzyme, whereas the charge density on the C4 atom of the pyridine ring is the major factor that governs the coenzyme activity.