Novel Approach For Early Detection Research Articles

Machine Learning Model for Anesthetic Risk Stratification for Gynecologic and Obstetric Patients: Cross-Sectional Study Outlining a Novel Approach for Early Detection.

Preoperative evaluation is important, and this study explored the application of machine learning methods for anesthetic risk classification and the evaluation of the contributions of various factors. To minimize the effects of confounding variables during model training, we used a homogenous group with similar physiological states and ages undergoing similar pelvic organ-related procedures not involving malignancies. Data on women of reproductive age (age 20-50 years) who underwent gestational or gynecological surgery between January 1, 2017, and December 31, 2021, were obtained from the National Taiwan University Hospital Integrated Medical Database. We first performed an exploratory analysis and selected key features. We then performed data preprocessing to acquire relevant features related to preoperative examination. To further enhance predictive performance, we used the log-likelihood ratio algorithm to generate comorbidity patterns. Finally, we input the processed features into the light gradient boosting machine (LightGBM) model for training and subsequent prediction. A total of 10,892 patients were included. Within this data set, 9893 patients were classified as having low anesthetic risk (American Society of Anesthesiologists physical status score of 1-2), and 999 patients were classified as having high anesthetic risk (American Society of Anesthesiologists physical status score of >2). The area under the receiver operating characteristic curve of the proposed model was 0.6831. By combining comorbidity information and clinical laboratory data, our methodology based on the LightGBM model provides more accurate predictions for anesthetic risk classification. Research Ethics Committee of the National Taiwan University Hospital 202204010RINB; https://www.ntuh.gov.tw/RECO/Index.action.

Neutrophil-targeted liposomal platform: a shift in novel approach for early detection and treatment of cancer metastasis

Tumor metastasis is responsible for 90% of cancer-associated deaths, and its early detection may decrease the likelihood of mortality. Studies have demonstrated that metastasis results from the interaction between “seeds” (tumor cells) and “soil” (pre-metastatic niche, PMN). As the first and most abundant immune cells to be recruited to PMN, neutrophils play a key role in the ultimate formation of metastatic foci through mechanisms such as supporting tumor cell growth, promoting angiogenesis, and shaping an immune-suppressive microenvironment. In this study, two distinct types of sialic acid (SA)-modified liposomes were prepared to target and regulate pro-metastatic neutrophils through the l-selectin receptor. One of these liposomes, named ICG@SAL, was used to encapsulate indocyanine green (ICG) and was specifically designed for the early detection of cancer metastasis. The other liposome, referred to as ABE/Cur@SAL, co-loaded abemaciclib (ABE) and curcumin (Cur), with the intention of suppressing the progression of metastatic tumor. Fluorescence imaging results from the mouse spontaneous metastasis model indicated that ICG@SAL demonstrated faster targeting and stronger accumulation in the metastatic organs than unmodified ICG liposomes (ICG@CL). This suggested that ICG@SAL could detect tumor metastasis at an early stage. The therapy with co-loaded liposomes in the mouse experimental lung metastasis model indicated that ABE/Cur@SAL could inhibit regulatory T (Treg) cell proliferation, enhance effector T cell activity and reduce tumorigenic factor release, implying that ABE/Cur@SAL could inhibit tumor metastasis. Overall, our work provided a sensitive and convenient approach to early diagnosis and treatment of tumor metastasis. ICG@SAL could be employed for the early detection of tumor metastasis, while ABE/Cur@SAL could be used to inhibit the development of tumor metastasis when early metastasis was identified.

Detection of critical cancer cells in human organs using dual demodulation photonic crystal fiber: Numerical study

This study reports a novel approach for early detection of malignant cancer cells in human organs using a birefringent photonic crystal fiber (PCF)-based optical sensor with dual demodulation. The PCF injects light into the middle hole, enhancing the radiated evanescent field. Analytes injected through the core cause a wavelength shift, measured by peak or dip shift. The proposed sensor has an optimal sensitivity of −7,940 nm/RIU, −8,265 nm/RIU, −9,747 nm/RIU, −9,006 nm/RIU, and −8,994 nm/RIU by peak shift and −8,745 nm/RIU, −10,728 nm/RIU, −8,721 nm/RIU, −10,113 nm/RIU, and −11,150 nm/RIU by dip shift for CRT-(Cervical tissue), BLD-(Blood), ADG-(Adrenal gland), BRT-(Breast) type-1 and type-2 cells, respectively. The PCF sensor presents a shorter sensing length of 550 µm. Results suggest the sensor is effective in detecting malignant neoplastic cells in earlier stages, making it a promising tool for cancer diagnosis and treatment.

An Early Chatter Detection Method Based on Multivariate Variational Mode Decomposition and Chatter Correlation Factor

Machining chatter is a self-excited vibration between the tool and the workpiece that affects surface quality and productivity. In this article, we propose a novel approach for early detection of machining chatter based on multivariate variational mode decomposition (MVMD) and chatter correlation factor (CCF). The proposed approach is featured by early chatter detection, and it is not susceptible to machining parameters. In order to obtain effective intrinsic mode functions (IMFs) and avoid mode mixing problem in MVMD, an automatic selection method of MVMD's parameters based on the difference of correlation coefficients (DCC) is developed. By using MVMD, multichannel signals are decomposed into multiple IMF groups with mode-alignment to extract common frequency components across all channel signals. Because multichannel signals are highly correlated at the chatter frequency when chatter occurs, the proposed CCF derived by correlation coefficients between the selected chatter-sensitive IMFs is sensitive to early chatter and can simultaneously measure the level of chatter. A support vector machine chatter identification model is obtained based on the CCF. The effectiveness of the proposed method is demonstrated by simulation and experimental findings.

A Novel Approach for Early Detection of Neovascular Glaucoma Using Fractal Geometry

A Novel Approach for Early Detection of Neovascular Glaucoma Using Fractal Geometry

A novel approach for early detection of Alzheimer’s disease using deep neural networks with magnetic resonance imaging

Since there is no cure for Alzheimer’s disease (AD), early detection is essential to mitigate the symptoms. However, early detection is extremely challenging because brain MRI scans during these initial stages look very similar to normal MRI scans to the human eye. To assist in early disease detection, we developed a convolution neural network deep learning model using a transfer learning approach to extract features and a custom dense layer to detect and classify the early stages of AD. We collected MRI scans from Kaggle and ADNI institutes at several stages of AD (mild, moderate, severe) as well as healthy control scans. We used Resnet-50, VGG-16, and DenseNet-169 base models and compared their performances for classifying the stages of AD. The test area under the curve (AUC) for each of the base models Resnet-50, VGG-16 and DenseNet-169 were 0.8334, 0.9047 and 0.8898, respectively, based on 1279 Kaggle MRI images and 0.7332, 0.7383 and 0.7133, respectively, based on 132 ADNI MRI images. VGG-16 outperformed both DenseNet-169 and Resnet-50 models and showed more accurate results. Our deep learning approach also detected and extracted the region of interest (ROI) as the superior frontal gyrus (SFG) and the hippocampus for the Kaggle and ADNI brain axial MRI images, respectively.

Kinematics approach with neural networks for early detection of sepsis (KANNEDS)

BackgroundSepsis is a severe illness that affects millions of people worldwide, and its early detection is critical for effective treatment outcomes. In recent years, researchers have used models to classify positive patients or identify the probability for sepsis using vital signs and other time-series variables as input.MethodsIn our study, we analyzed patients’ conditions by their kinematics position, velocity, and acceleration, in a six-dimensional space defined by six vital signs. The patient is affected by the disease after a period if the position gets “near” to a calculated sepsis position in space. We imputed these kinematics features as explanatory variables of long short-term memory (LSTM), convolutional neural network (CNN) and linear neural network (LNN) and compared the prediction accuracies with only the vital signs as input. The dataset used contained information of approximately 4800 patients, each with 48 hourly registers.ResultsWe demonstrated that the kinematics features models had an improved performance compared with vital signs models. The kinematics features model of LSTM achieved the best accuracy, 0.803, which was nine points higher than the vital signs model. Although with lesser accuracies, the kinematics features models of the CNN and LNN showed better performances than vital signs models.ConclusionApplying our novel approach for early detection of sepsis using neural networks will prove to be an invaluable, more accurate method than considering only simple vital signs as input variables. We expect that other researchers with similar objectives can use the model presented in this innovative approach to improve their results.

Inter-Collateral Diabetic Retinopathy Extraction and Evaluation Using Trained Datasets of Multiple Feature Set

Diabetic Retinopathy (DR) is caused due to un-mounted diabetic comorbidities. The patients suffer complete vision blindness if untreated or diagnosed on later stage. In this article, we propose a novel approach for early detection and prediction using trained datasets of multiple features. The process expansion is resultant of multiple stage attribute extraction via a series of inter-collateral parameters of diabetics. Typically, the proposed technique is designed and developed on a multi-value and multi-dimension datasets such as comorbidities history of patient encountered during diabetics. The proposed technique uses collateral attributes in evaluating retinopathy status and thereby validates the extracted DR under threshold value comparisons. The results are computed using HADOOP framework for recursive pattern and feature evaluation. The trial is processed on UCL digital library datasets with estimated performance of 98.7% with extraction and 92.34% with value True-Positive (TP) prediction.

Multi-input convolutional neural network for breast cancer detection using thermal images and clinical data

Background and objectiveBreast cancer is the most common cancer in women. While mammography is the most widely used screening technique for the early detection of this disease, it has several disadvantages such as radiation exposure or high economic cost. Recently, multiple authors studied the ability of machine learning algorithms for early diagnosis of breast cancer using thermal images, showing that thermography can be considered as a complementary test to mammography, or even as a primary test under certain circumstances. Moreover, although some personal and clinical data are considered risk factors of breast cancer, none of these works considered that information jointly with thermal images. MethodsWe propose a novel approach for early detection of breast cancer combining thermal images of different views with personal and clinical data, building a multi-input classification model which exploits the benefits of convolutional neural networks for image analysis. First, we searched for structures using only thermal images. Next, we added the clinical data as a new branch of each of these structures, aiming to improve its performance. ResultsWe applied our method to the most widely used public database of breast thermal images, the Database for Mastology Research with Infrared Image. The best model achieves a 97% accuracy and an area under the ROC curve of 0.99, with a specificity of 100% and a sensitivity of 83%. ConclusionsAfter studying the impact of thermal images and personal and clinical data on multi-input convolutional neural networks for breast cancer diagnosis, we conclude that: (1) adding the lateral views to the front view improves the performance of the classification model, and (2) including personal and clinical data helps the model to recognize sick patients.

DEODORANT: a novel approach for early detection and prevention of polycystic ovary syndrome using association rule in hypergraph with the dominating set property

Present online health discussion forums generate the bountiful amount of digitized data through health-blogs, posts, tweets, and chats in the social media. People post queries, health issues they undergo along with the symptoms, diagnosis, and clinical reports to get direction for preventive measures and medical relief. As a case study, this work focus on detection of Polycystic Ovary Syndrome, a prevalent condition that affects a woman’s hormone levels. This PCOS problem has been investigated as it forms high-risk factor for infertility, heart disease, diabetes, stroke and many such diseases. We propose a novel model named as DEODORANT (Detection and prEvention of polycystic Ovary synDrome using assOciation rule hypeRgrAph and domiNating set properTy) to derive prospective use of real-time mining data. The unstructured data collected from various media sources are preprocessed using NLTK and association rules are derived by applying apriori algorithm. These association rules are represented in hypergraph and then regenerated as line graph to make it suitable for cluster construction. Spectral clustering is performed on line graph to partition into clusters of hypergraphs. By applying dominating set property on the resultant hypergraph, required inferences can be elicited. From the experimental results, support value of the outcome derived from the dominating set of each cluster has exhibited the symptoms and the causes with percentage ranking. It is evident that they get aligned with precise result portraying real statistics. This type of analysis will empower doctors and health organizations to keep track of the diseases, their symptoms for early detection and safe recovery.

Early Sepsis Detection with Deep Learning on EHR Event Sequences

Background: Sepsis is a clinical condition involving an extreme inflammatory response to an infection, and is associated with high morbidity and mortality. Without intervention, this response can progress to septic shock, organ failure and death. Every hour that treatment is delayed mortality increases. Early identification of sepsis is therefore important for a positive outcome.
 Methods: We constructed predictive models for sepsis detection and performed a register-based cohort study on patients from four Danish municipalities. We used event-sequences of raw electronic health record (EHR) data from 2013 to 2017, where each event consists of three elements: a timestamp, an event category (e.g. medication code), and a value. In total, we consider 25.622 positive (SIRS criteria) sequences and 25.622 negative sequences with a total of 112 million events distributed across 64 different hospital units. The number of potential predictor variables in raw EHR data easily exceeds 10.000 and can be challenging for predictive modeling due to this large volume of sparse, heterogeneous events. Traditional approaches have dealt with this complexity by curating a limited number of variables of importance; a labor-intensive process that may discard a vast majority of information. In contrast, we consider a deep learning system constructed as a combination of a convolutional neural network (CNN) and long short-term memory (LSTM) network. Importantly, our system learns representations of the key factors and interactions from the raw event sequence data itself.
 Results: Our model predicts sepsis with an AUROC score of 0.8678, at 11 hours before actual treatment was started, outperforming all currently deployed approaches. At other prediction times, the model yields following AUROC scores. 15 min: 0.9058, 3 hours: 0.8803, 24 hours: 0.8073.
 Conclusion: We have presented a novel approach for early detection of sepsis that has more true positives and fewer false negatives than existing alarm systems without introducing domain knowledge into the model. Importantly, the model does not require changes in the daily workflow of healthcare professionals at hospitals, as the model is based on data that is routinely captured in the EHR. This also enables real-time prediction, as healthcare professionals enters the raw events in the EHR.

Abstract P4-02-10: Breast cancer surveillance in high-risk women with dynamic contrast-enhanced magnetic resonance imaging every 6 months: Results from a single institution study

Abstract Purpose: To develop a novel approach for early detection of breast cancer and examine molecular features of screen detected cancers in prospectively ascertained high-risk women undergoing semi-annual dynamic contrast-enhanced breast magnetic resonance imaging (DCE-MRI) for women at high genetic risk. Background: Women with a personal or family history of breast cancer and genetic mutation carriers of BRCA1 and BRCA2 have a higher than normal risk of breast cancer. An intensified screening surveillance regimen is an early detection strategy in high-risk women. The American Cancer Society recommends annual DCE-MRI in addition to annual mammogram based off several pivotal screening studies that demonstrated improved sensitivity and cancer detection rates and decreased interval cancer rates with the addition of annual DCE-MRI. Questions remain regarding the optimal screening modality and interval regimen in these high-risk women. Methods: Between 2004 and 2016, we assembled a prospective cohort of high-risk women undergoing semi-annual DCE-MRI and annual mammography. To be eligible, women had a lifetime breast cancer risk >20% and/or tested positive for a pathogenic mutation using a cancer gene panel including BRCA1, BRCA2, CDH1, PALB2, CHEK2 and other cancer susceptibility genes in the DNA repair pathway. Somatic mutation events in screen-detected tumors were investigated using UW-OncoPlex cancer gene panel using DNA extracted from FFPE shavings. Results: 295 women were recruited to the study; 44% of the study participants had pathogenic mutations in BRCA1 or BRCA2 genes. At a median follow-up of 3.3 years (range 0-12 years), 3 DCIS and 13 early stage invasive breast cancers were detected, of which 14 occurred in subjects with identifiable pathogenic mutations (11 BRCA1, 2 BRCA2, 1 CDH1). The incidence rate is 1.3% in all subjects, but 3.5 % per year in BRCA1 carriers. DCE-MRI identified all 13 invasive cancers at a mean size of 0.61 cm (range 0.1-1.0 cm); none had lymph node metastasis. No interval cancers occurred. In addition, 7 of the breast cancers were detected on DCE-MRI imaging obtained at the 6 months screening interval; they would be interval cancers if only annual screening were implemented. There was very little DNA for somatic mutation testing in the majority of cases. However, as expected, there was heterogeneity in the spectrum of mutations but the most commonly somatically mutated gene in the early cancers was TP53. Conclusions: DCE-MRI every 6 months performed well for early detection of invasive breast cancer in high-risk women, accomplishing the ultimate goal of breast cancer screening—detecting node-negative, invasive tumors less than 1 cm. Semi-annual DCE-MRI performed especially well in BRCA1 mutation carriers at risk for the most aggressive subtype of breast cancer. Further interventional studies evaluating this novel screening approach are warranted to personalize breast cancer risk assessment and prevention. Citation Format: Whitaker K, Guindalini R, Abe H, Sheeth D, Huo D, Hong S, Churpek J, Verp M, Obeid E, Zheng Y, Amico A, Yoshimatsu T, Olopade O. Breast cancer surveillance in high-risk women with dynamic contrast-enhanced magnetic resonance imaging every 6 months: Results from a single institution study [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P4-02-10.

Easy Screening for Mild Alzheimer's Disease and Mild Cognitive Impairment from Elderly Speech.

This study presents a novel approach for early detection of cognitive impairment in the elderly. The approach incorporates the use of speech sound analysis, multivariate statistics, and data-mining techniques. We have developed a speech prosody-based cognitive impairment rating (SPCIR) that can distinguish between cognitively normal controls and elderly people with mild Alzheimer's disease (mAD) or mild cognitive impairment (MCI) using prosodic signals extracted from elderly speech while administering a questionnaire. Two hundred and seventy-three Japanese subjects (73 males and 200 females between the ages of 65 and 96) participated in this study. The authors collected speech sounds from segments of dialogue during a revised Hasegawa's dementia scale (HDS-R) examination and talking about topics related to hometown, childhood, and school. The segments correspond to speech sounds from answers to questions regarding birthdate (T1), the name of the subject's elementary school (T2), time orientation (Q2), and repetition of three-digit numbers backward (Q6). As many prosodic features as possible were extracted from each of the speech sounds, including fundamental frequency, formant, and intensity features and mel-frequency cepstral coefficients. They were refined using principal component analysis and/or feature selection. The authors calculated an SPCIR using multiple linear regression analysis. In addition, this study proposes a binary discrimination model of SPCIR using multivariate logistic regression and model selection with receiver operating characteristic curve analysis and reports on the sensitivity and specificity of SPCIR for diagnosis (control vs. MCI/mAD). The study also reports discriminative performances well, thereby suggesting that the proposed approach might be an effective tool for screening the elderly for mAD and MCI.

Brain-Specific Serum Biomarkers Predict Neurological Morbidity in Diagnostically Diverse Pediatric Intensive Care Unit Patients.

Unexpected neurological morbidity in Pediatric Intensive Care Units (PICUs) remains high and is difficult to detect proactively. Brain-specific biomarkers represent a novel approach for early detection of neurological injury. We sought to determine whether serum concentrations of neuron-specific enolase (NSE), myelin basic protein (MBP), and S100B, specific for neurons, oligodendrocytes, and glia, respectively, were predictive of neurological morbidity in critically ill children. Serum was prospectively collected on days 1-7 from diagnostically diverse PICU patients (n=103). Unfavorable neurological outcome at hospital discharge was defined as Pediatric Cerebral Performance Category (PCPC) score of 3-6 with a deterioration from baseline. NSE, MBP, and S100B concentrations were measured by enzyme-linked immunosorbent assay. Peak biomarker levels were greater in patients with unfavorable versus favorable neurological outcome [NSE 39.4±44.1 vs. 12.2±22.9ng/ml (P=0.005), MBP 9.1±11.5 vs. 0.6±1.3ng/ml (P=0.003), S100B 130±232 vs. 34±70pg/ml (P=0.04), respectively; mean±SD]. Peak levels were each independently associated with unfavorable neurological outcome when controlling for presence of primary neurologic admission diagnosis and poor baseline PCPC using logistic regression analysis (NSE, P=0.04; MBP, P=0.004; S100B, P=0.04), and had the following receiver operating characteristics: NSE 0.75 (0.58, 0.92), MBP 0.81 (0.66, 0.94), and S100B 0.80 (0.67, 0.93) (area under the curve [95% confidence intervals]). Prospectively collected brain-specific serum biomarkers predict unfavorable neurological outcome in critically ill children. Serum biomarkers used in conjunction with clinical data could be used to generate models predicting early detection of neurological injury, allowing for more timely diagnostic and therapeutic interventions, potentially reducing neurological morbidity in the PICU.

A novel approach for early detection of impending voltage collapse events based on the support vector machine

Summary This paper proposes an approach to detect the possibility of long-term voltage instability, based on online measurement of system bus voltages. An optimization framework is proposed to determine the maximum loading points, with different load increase patterns and different levels of reactive power output. The operating conditions so obtained are used as the training database for an artificial intelligence classifier based on the support vector machines. In an online application, the support vector machine classifier helps in detecting the probability of some generators operating at high reactive power output, which is an important indicator of an impending voltage collapse. The proposed framework is tested with the IEEE 39 bus and the Nordic 32 bus systems. The test results demonstrate that the proposed scheme gives reliable prediction of the power system long-term voltage stability.

Accurate Sensitivity of Quantum Dots for Detection of HER2 Expression in Breast Cancer Cells and Tissues

Here we introduce novel optical properties and accurate sensitivity of Quantum dot (QD)-based detection system for tracking the breast cancer marker, HER2. QD525 was used to detect HER2 using home-made HER2-specific monoclonal antibodies in fixed and living HER2(+) SKBR-3 cell line and breast cancer tissues. Additionally, we compared fluorescence intensity (FI), photostability and staining index (SI) of QD525 signals at different exposure times and two excitation wavelengths with those of the conventional organic dye, FITC. Labeling signals of QD525 in both fixed and living breast cancer cells and tissue preparations were found to be significantly higher than those of FITC at 460-495nm excitation wavelengths. Interestingly, when excited at 330-385nm, the superiority of QD525 was more highlighted with at least 4-5 fold higher FI and SI compared to FITC. Moreover, QDs exhibited exceptional photostability during continuous illumination of cancerous cells and tissues, while FITC signal faded very quickly. QDs can be used as sensitive reporters for in situ detection of tumor markers which in turn could be viewed as a novel approach for early detection of cancers. To take comprehensive advantage of QDs, it is necessary that their optimal excitation wavelength is employed.

"Volatile signature of melanoma - a novel approach for early detection"

"Volatile signature of melanoma - a novel approach for early detection"

Detection and Quantification of the Evolution Dynamics of Apoptosis Using the PET Voltage Sensor 18F-Fluorobenzyl Triphenyl Phosphonium

Apoptosis is a key mechanism in numerous pathologies. However, there are no effective noninvasive means available for an early detection and quantitative assessment of evolution dynamics of the apoptotic process. Here, we have characterized the ability of the novel PET voltage sensor (18)F-fluorobenzyl triphenyl phosphonium ((18)F-FBnTP) to quantify the time-dependent apoptotic action of the taxanes paclitaxel and docetaxel in vitro and in vivo. The duration-dependent treatment effect of paclitaxel on (18)F-FBnTP uptake was assayed in human MDA-MB-231 breast carcinoma cells. The expression of the proapoptotic Bax and antiapoptotic Bcl-2 mitochondrial proteins, release of the apoptogen cytochrome c, and activation of executioner caspase-3 were determined by Western blotting. The fraction of viable cells was determined using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide. The effect of docetaxel on (18)F-FBnTP and (18)F-FDG uptake in orthotopic prostate tumors in mice was compared. (18)F-FBnTP cellular uptake in viable cells declined linearly with the increasing duration of paclitaxel treatment, from 3 to 24 h, and plateaued at 48 h. The extent of decrease of (18)F-FBnTP correlated strongly with the Bax-to-Bcl-2 ratio (R(2) = 0.83) and release of cytochrome c (R(2) = 0.92), but preceded in time the caspase-3 cleavage. The P-glycoprotein blocker verapamil did not interfere with (18)F-FBnTP cellular uptake. (18)F-FBnTP prostate tumor contrast was greater than (18)F-FDG prostate tumor contrast. Docetaxel caused a marked decrease (52.4%) of (18)F-FBnTP tumor uptake, within 48 h, whereas (18)F-FDG was much less affected (12%). The voltage sensor (18)F-FBnTP is a viable means for quantification of paclitaxel pharmacodynamics. (18)F-FBnTP permits the detection of paclitaxel apoptotic action in vivo earlier than does (18)F-FDG. (18)F-FBnTP may afford a novel approach for early detection and quantitative assessment of the cumulative-effect kinetics of proapoptotic drugs and conditions using PET.

Sensitivity and specificity of faecal tumour M2 pyruvate kinase for detection of colorectal adenomas in a large screening study

The measurement of faecal tumour M2 pyruvate kinase (tumour M2 PK) has been proposed as a novel approach for early detection of colorectal cancer (CRC). However, as regards the potential of the test to detect precursors to CRC, an issue that is highly relevant to estimate its use in reducing CRC incidence and mortality, the available evidence is scant and controversial. The aim of our study was to determine the performance characteristics of the tumour M2 PK test with respect to colorectal adenomas in the target population of screening. Among 1082 participants of screening colonoscopy in Germany, of whom 30% had any adenoma and 10% had an advanced adenoma, the median (interquartile range) tumour M2 PK level in the whole study population was 1.3 U ml−1 (0.3–3.3). At a cutoff value of 4 U ml−1, sensitivity was 22 and 23% for detection of advanced and other adenomas, respectively, whereas specificity was 82%. The area under the receiver-operating characteristics curve (95% confidence interval) was 0.54 (0.51–0.58) and 0.56 (0.52–0.59) for advanced and other adenomas, respectively. In conclusion, the tumour M2 PK test has only very limited potential to distinguish between people bearing precursors to CRC and people with no finding at colonoscopy.