Artificial intelligence-based automated preprocessing and classification of impacted maxillary canines in panoramic radiographs.

OBJECTIVES: Automating the digital workflow for diagnosing impacted canines using panoramic radiographs (PRs) is challenging. This study explored feature extraction, automated cropping, and classification of impacted and nonimpacted canines as a first step. METHODS: A convolutional neural network with SqueezeNet architecture was first trained to classify two groups of PRs (91with and 91without impacted canines) on the MATLAB programming platform. Based on results, the need to crop the PRs was realized. Next, artificial intelligence (AI) detectors were trained to identify specific landmarks (maxillary central incisors, lateral incisors, canines, bicuspids, nasal area, and the mandibular ramus) on the PRs. Landmarks were then explored to guide cropping of the PRs. Finally, improvements in classification of automatically cropped PRs were studied. RESULTS: Without cropping, the area under the curve (AUC) of the receiver operating characteristic (ROC) curve for classifying impacted and nonimpacted canine was 84%. Landmark training showed that detectors could correctly identify upper central incisors and the ramus in ∼98% of PRs. The combined use of the mandibular ramus and maxillary central incisors as guides for cropping yielded the best results (∼10% incorrect cropping). When automatically cropped PRs were used, the AUC-ROC improved to 96%. CONCLUSIONS: AI algorithms can be automated to preprocess PRs and improve the identification of impacted canines.

Fourier Transform Infrared Spectroscopy (FT-IR) of Pseudomonas aeruginosa post photodynamic therapy with Curcumin in vitro.

Alternative therapies against pathogens are under intense investigation because of their increasing resistance to antibiotics. Photodynamic therapy (PDT) is one such alternative that has shown promising results. However, for the widespread use of PDT, it is essential to decipher underlying mechanisms, so as to improve PDT's therapeutic applications. Because of this, we have studied biochemical changes in pathogen Pseudomonas aeruginosa, a medically important bacteria that has developed antibiotic resistance, after PDT with curcumin photosensitizer. Results show a drastic decrease in α-helix protein and increased disordered and ß-sheet secondary structure proteins in P. Aeruginosa post-PDT compared to control. Interestingly, these biochemical changes differ from PDT of pathogens Leishmania braziliensis and Leishmania major with photosensitizer methylene blue. This observation underlines the need for extensive studies on PDT of different pathogens to understand mechanisms of action and develop better PDT strategies.

Raman spectroscopy of healthy, injured and amniotic membrane treated rat spinal cords.

Spinal cord injury is a significant public health issue with high psychological and financial costs to both the family and the society. Effective treatment strategies are hence of immense value. Several reports have suggested application of amniotic membrane for treating injuries, and there is evidence that it may be used to treat spinal injuries. In this animal model study, we explore biochemical changes in amniotic membrane treated injured spinal cord with respect to untreated injured and uninjured spinal cord using Raman spectroscopy. Multivariate statistical analysis is able to classify control, untreated, and treated with 92%, 87%, and 80% efficiency, respectively; suggesting unique biochemical changes in each group. Such studies may lead to development of minimally invasive methodologies for spinal cord injury treatment monitoring.

PCR-RFLP and FTIR-based detection of high-risk human papilloma virus for cervical cancer screening and prevention.

BACKGROUND: Approximately 70% of cervical carcinoma cases show the presence of high-risk Human Papilloma Virus (HPV), especially HPV-16 and HPV-18, and can be used to stratify high risk patients from low risk and healthy. Currently, molecular biology techniques such as polymerase chain reaction (PCR) are used to identify the presence of virus in patient samples. While the methodology is highly sensitive, it is labor intensive and time-consuming. Alternative techniques, such as vibrational spectroscopy, has been suggested as a possible rapid alternative. Therefore, in this study, we evaluate the efficiency of cervical fluid Fourier Transform Infrared spectroscopy (FTIR) in patient risk stratification informed by PCR. METHODS: Cervical fluid samples (n = 91) were obtained from patients who have undergone routine Papanicolaou (Pap) test. Viral genome was identified and classified as high/low-risk by PCR-Restriction Fragment Length Polymorphism (PCR-RFLP). FTIR spectra were acquired from samples identified by PCR-RFLP as No-HPV (n = 10), high-risk HPV (n = 7), and low-risk HPV (n = 7). RESULTS: Of the 91 samples, was detected the viral genome by PCR in 36 samples. Of these 36 samples, nine samples were identified to contain high-risk HPV (HR-HPV) and nine samples were found to have low-risk HPV (LR-HPV). The FTIR spectra acquired from No-HPV, LR-HPV, and HR-HPV showed differences in 1069, 1437, 1555, 1647, 2840, 2919, and 3287 cm-1 bands. Principal Component Analysis (PCA) showed distinct clusters for No-HPV and HR-HPV and No-HPV and LR-HPV, but there was significant overlap in the clusters of HR-HPV and LR-HPV. PCA-Linear Discriminant Analysis (PC-LDA) after Leave One Out Cross Validation (LOOCV) classified No-HPV from HR-HPV and No-HPV from LR-HPV with 100% efficiency in the 1400-1800 cm-1 spectral range. LOOCV classifications for LR-HPV and HR-HPV from each other were 71 and 75%, respectively, in the 2800-3400 cm-1 spectral range. CONCLUSIONS: The results highlight the high sensitivity of PCR-RFLP in HPV identification and show that FTIR can classify samples identified as healthy, low, and high-risk samples by PCR-RFLP. GENERAL SIGNIFICANCE: We show the possibility of using FTIR for initial cervical cancer risk stratification followed by detailed PCR-RFLP investigations for suspect cases.

Infrared Spectroscopy Based Study of Biochemical Changes in Saliva during Maximal Progressive Test in Athletes.

This study aims to explore biochemical changes in saliva during cardiorespiratory exercise using attenuated-total-reflectance-Fourier-transform-infrared-spectroscopy (ATR-FTIR). Saliva and blood samples were obtained from six athletes at rest, and after running at speeds of 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, and 20 kilometers-per-hour (km/h) on a treadmill (maximal stress test). Saliva ATR-FTIR spectra were analyzed using deconvolution and multivariate analysis. Area-under-the-curve calculations suggest differential changes in glucose, lactate, protein, lipids, carbohydrate and phosphate content in saliva during the test. Increases in glucose and lactate levels with increasing speeds were verified by simultaneous measurement of blood glucose and lactate levels using standard equipment (Roche®). Multivariate principal-component-analysis (PCA) showed discrete clusters for low (rest-14 km/h) and high (15 - 20 km/h) speeds, and PCA-linear-discriminant-analysis showed 100% classification of 18 - 20 km/h as high speed. Overall, results suggest the possibility of using this non-invasive saliva-based ATR-FTIR method for biochemical assessment during sports exercise and stress tests.

Biochemical changes in Leishmania braziliensis after photodynamic therapy with methylene blue assessed by the Fourier transform infrared spectroscopy.

Photodynamic therapy (PDT) with photosensitizer methylene blue was applied to Leishmania braziliensis, and Fourier transform infrared (FTIR) spectroscopy was used to study biochemical changes in the parasite after PDT in comparison to untreated (C), only irradiation (I), and only photosensitizer (PS). Spectral analysis suggests increase in lipids, proteins, and protein secondary structures in PDT compared with C and decrease in nucleic acids and carbohydrates. Interestingly, these trends are different from PDT of Leishmania major species, wherein lipids decrease; there are minimal changes in secondary structures and increase in nucleic acids and carbohydrates. The study thus suggests possibility of different biomolecular players/pathways in PDT-induced death of L. braziliensis and L. major.

Saliva Preparation Method Exploration for ATR-FTIR Spectroscopy: Towards Bio-fluid Based Disease Diagnosis.

Saliva has garnered a lot of interest as a non-invasive, easy to collect, and biochemical rich sample for attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR) based disease diagnosis. Although a large number of studies have explored its potential, the preparation methods used differ greatly. For large scale clinical studies to aid translation into clinics, the collection/processing methodology needs to be standardized. Therefore, in this study, we explored different saliva collection (spitting, method A/cotton soaking, method B) and processing protocols (unprepared, TS; supernatant from the centrifugation, CS; and drying, C) to find which gives the best ATR-FTIR signals. Analysis showed highest proteins, carbohydrates, amino acids, and nucleic acid + proteins/lipids in BTS, BCS, ACS, and BC, respectively. Notably, only BC shows a 1377 cm-1 nucleic acid band that is also uniquely identified in multivariate analysis. We conclude that the collection-processing protocol should be based on a biochemical component that best gives a differential diagnosis.

Raman spectroscopy-multivariate analysis related to morphological surface features on nanomaterials applied for dentin coverage.

Raman spectroscopy and scanning electron microscopy (SEM) were used to investigate the effect of coating materials and acidulated phosphate fluoride gel (APF) treatment on dentin before and after erosion-abrasion cycles. A multi-walled carbon nanotube/graphene oxide hybrid carbon-based material (MWCNTO-GO), nanohydroxyapatite (nHAp), or a combined composite (nHAp/MWCNTO-GO) were used as a coating. Seventy root dentin fragments obtained from 40 bovine teeth were prepared and divided into groups (n = 10): negative control, artificial saliva - C, positive control - APF; nHAp; MWCNTO-GO; APF_nHAp; APF_MWCNTO-GO and APF_nHAp/MWCNTO-GO. All samples were subjected to cycles of demineralization (orange juice, pH ~3.7, room temperature, 1 min) followed by remineralization (saliva, 37 °C, 1 h). The remineralization procedures were followed by tooth brushing (150 strokes). The above cycle was repeated 3×/day for 5 days. The previous APF treatment of dentin allowed a better affinity of nHAp and MWCNTO-GO with the inorganic and organic portion of dentin, respectively. This interaction indicates the formation of a protective layer for the dentin surface and for the collagen giving possible protection against erosion. SEM micrographs illustrated the formation of a protective layer after application of the biomaterials and that it was partially or totally removed after the erosion and abrasion. Raman spectroscopy combined with multivariate analysis could distinguish samples with respect to treatment efficacy. The APF_nHAP/MWCNT-GO composite has shown to be a promising material since it has binding characteristics both to the inorganic and organic portion of the dentin and reduced solubility. Mineral-to-matrix ratio (MMR) parameter analysis confirmed the binding capability of MWCNTO-GO-based materials to dentin.

ATR-FTIR spectroscopy and CDKN1C gene expression in the prediction of lymph nodes metastases in papillary thyroid carcinoma.

Thyroid cancer has become in recent years the most common endocrine malignancy. Among its different types, papillary thyroid carcinoma (PTC) has the highest incidence. PTC is slow growing, but shows a high rate of lymph node metastasis. Tissue biochemical characterization and identification of molecular markers can facilitate stratification of patients into those requiring surgical assessment of lymph nodes and patients for whom this surgical procedure is unnecessary; thus, leading to a more accurate prognosis. To this end, the study aimed to predict lymph node metastasis by Attenuated Total Reflectance - Fourier transform infrared (ATR-FTIR) spectroscopy of primary PTC tumors. Another objective of the study was to determine whether CCNA1, CDKN1C, FOS, HSPA5, JUN, KSR1, MAP2K6, MAPK8IP2 and SFN gene expression in primary PTC tumors could be used as predictive markers of lymph node metastasis. Three PTC with lymph node involvement (PTC+), six PTC without lymph node involvement (PTC-), and five normal (N) thyroid tissues were used for FTIR spectroscopy analysis; while 18 PTC+, 17 PTC-, and 6 N samples were used for molecular analysis by real-time quantitative PCR (RT-qPCR). FTIR spectral analysis revealed changes in phosphate groups possibly associated with nucleic acid (1236 cm-1), and protein/lipids (1452, 2924, 3821 cm-1) in PTC + compared to PTC-, and multivariate analysis could distinguish the two groups. Molecular analysis showed significant increase in CDKN1C gene expression in PTC + compared to PTC-. Being a cell growth regulator, increased CDKN1C provides some supporting evidence to the FTIR spectroscopy based finding of increased nucleic acids in PTC+. Thus, the study suggests the possibility of using FTIR spectroscopy and CDKN1C expression for predicting metastasis using primary tumor alone.

Adverse effects of respiratory disease medicaments and tooth brushing on teeth: A scanning electron microscopy, X-ray fluorescence and infrared spectroscopy study.

The present study aims to evaluate the effect of brushing with fluoride dentifrice on teeth severely affected by erosion due to respiratory medicaments. Enamel (n = 50) and dentin (n = 50) bovine specimens were prepared and treated with artificial saliva (S-control), acebrofilin hydrochloride (AC), ambroxol hydrochloride (AM), bromhexine hydrochloride (BR), and salbutamol sulfate (SS) and subjected to cycles of demineralization (immersing in 3 mL, 1 min, three times a day at intervals of 1 hr, for 5 days) followed by remineralization (saliva, 37°C, 1 hr). Simulated brushing with fluoridated toothpaste was performed using 810 strokes in a reciprocal-action brushing simulator. Scanning electron microscopy, micro energy dispersive X-ray fluorescence (µ-EDXRF) spectroscopy and attenuated total reflection Fourier transform infrared (ATR FTIR) spectroscopy were then performed. µ-EDXRF images showed extensive erosion after treatment with all medicaments. SEM images showed enamel erosion in order SS > BR > AC = AM > S after brushing and fluoridation. FTIR results were in agreement. In case of dentin, µ-EDXRF measurements showed significant difference in mineral content (percent weight of calcium and phosphate) in SS + brushing + fluoridation treated enamel compared to control, while µ-EDXRF images showed erosive effects in the order SS > AM>BR > AC = S post brushing + fluoridation. SEM images showed erosion in the order SS > AM = BR > AC > S post brushing + fluoridation. Again, FTIR multivariate results were in agreement. Overall, our study shows that proper oral care is critical when taking certain medication. The study also demonstrates the possible use of FTIR for rapid clinical monitoring of tooth erosion in clinics.

In vivo Raman spectroscopic characteristics of different sites of the oral mucosa in healthy volunteers.

OBJECTIVES: Investigate the biochemistry of in vivo healthy oral tissues through Raman spectroscopy. We aimed to characterize the biochemical features of healthy condition in oral subsites (buccal mucosa, lip, tongue, and gingiva) of healthy subjects. More specifically, we investigated Raman spectral characteristics and biochemical content of in vivo healthy tissues on Brazilian population. This characterization can be used to better define normal tissue and improve the detection of oral premalignant conditions in future studies. MATERIALS AND METHODS: For spectroscopic analysis a Raman spectrometer (Kaiser Optical Systems imaging spectrograph Holospec, f / 1.8i-NIR) coupled with a laser 785 nm, 60 mW was used. Raman measurements were obtained by means of an optical fiber (EMVision fiber optic probe) coupled between the laser and the spectrometer. Three spectra per site were acquired from the lip, buccal mucosa, tongue, and gingiva of ten healthy volunteers. This resulted in 30 spectra per oral sub-site and in total 120 spectra. RESULTS: We report detailed biochemical information on these subsites and their relative composition based on deconvolution studies of their spectra. Finally, we also report classification efficiency of 61, 83, 41, and 93% for buccal, gingiva, lip, and tongue respectively after applying multivariate statistical tools. CONCLUSIONS: We quantitated the contribution of various biochemicals in terms of percentage, and this will enable comparison not only across anatomical sites but also across studies. Raman spectroscopy can rapidly probe tissue biochemistry of healthy oral regions. Moreover, the study suggests the possibility of using Raman spectroscopy combined with signal processing and multivariate analysis methods to differentiate the oral sites in healthy conditions and compare with pathological conditions in future studies. CLINICAL RELEVANCE: The spectral characterization of the healthy condition of oral tissues by a noninvasive, label-free, and real-time analytical techniques is important to create a spectral reference for future diagnosis of pathological conditions.

FTIR study of Achilles tendinopathy: protein secondary structure changes in tendon post injury

Abstract Introduction Tendinopathy, a disease involving tendon inflammation and damage, results in considerable productivity and economic losses for the patient and the society. Currently available diagnosis/ treatment monitoring modalities are less efficacious and highly subjective, underlying the need for better tools. Previously, we have shown that Fourier Transform Infrared (FTIR) spectroscopy has potential in tendinopathy diagnosis/ monitoring, when combined with multivariate statistical analysis. The current study explores the spectral range that gave the best results in statistical analysis, to discover protein secondary structures associated with tendon injury and recovery that can act as markers of disease. Methods Animals (n=60) underwent a surgery in which Achilles tendon were injured by dropping a 20g weight. Rats were divided into three groups (n=20) - control (C), Achilles tendon injury (I) and Achilles tendon injury treated with amniotic membrane fragment treated (T). FTIR spectra were obtained from each group 3, 7, 14, and 28 days post injury/ treatment. Results Triple helix, β-turn, and disordered structure levels differ between control, injured and treated tendons over the time period studied. Parallel β-sheets increase steadily over time in treated tendons compared to control and injured. Conclusion Combined analysis of Triple helix, β- sheets, β-turn, and disordered structure levels may be useful for tendinopathy diagnosis and treatment monitoring. However, further studies in this area are required to confirm the findings.

Erratum: Raman spectral post-processing for oral tissue discrimination - a step for an automatized diagnostic system: erratum.

[This corrects the article on p. 5218 in vol. 8, PMID: 29188115.].

Human papillomavirus detection using PCR and ATR-FTIR for cervical cancer screening.

The human papillomavirus (HPV) genital infection is considered one of the most common sexually transmitted diseases worldwide, and has been associated with cervical cancer. The objective of this study was to investigate the efficacy of the diagnostic methods: polymerase chain reaction (PCR) and Fourier transform infrared (FTIR) equipped with an ATR (Attenuated Total Reflectance) unit (Pike Tech) spectroscopy, to diagnose HPV infection in women undergoing gynecological examination. Seventeen patients (41.46%) of the 41 patients analyzed were diagnosed with exophytic/condyloma acuminate lesions by clinical analysis, 29 patients (70.7%) (G1 group) of the 41 patients, showed positive result for HPV cell injury by oncotic colpocitology and 12 patients (29.3%) (G2 group), presented negative result for cellular lesion and absence of clinical HPV lesion. Four samples were obtained per patient, which were submitted oncotic colpocitology analysis (Papanicolau staining, two samples), PCR (one sample) and ATR-FTIR analysis (one sample). L1 gene was amplified by PCR technique with specific GP5+/GP6+ and MY09/MY11 primers. PCR results were uniformly positive for presence of HPV in all analyzed samples. Multivariate analysis of ATR-FTIR spectra suggests no significant biochemical changes between groups and no clustering formed, concurring with results of PCR. This study suggests that PCR and ATR-FTIR are highly sensitive technique for HPV detection.

Tendinopathy diagnosis and treatment monitoring using attenuated total reflectance-Fourier transform infrared spectroscopy.

Tendinopathy, an important sports injury afflicting athletes and general public, is associated with huge economic losses. The currently used diagnostic tests are subjective, show moderate sensitivity and specificity; while treatment failures persist despite advances in therapy. This highlights the need for tendinopathy diagnostic and treatment monitoring tools. This study investigates tendon injury, natural healing and effect of treatment using ATR-FTIR complemented with histopathology. Control (C), injured (I) and treated (T) rat tendons were extracted 3, 7, 14 and 28 days post-injury/treatment, representing phases of healing; and subjected to hematoxylin & eosin staining as well as spectroscopy. While C showed no change, I- and T-related histological changes could be clearly observed in stained sections. ATR-FTIR spectra highlighted the biochemical changes within groups. Multivariate analysis could classify C, I and T with 75%; different days between groups with 84%; and different days within group with 65% efficiency. Results suggest that such analysis can not only identify C, I or T but also different phases of healing. Difference between I and T at different time points also suggest change in rate of healing. Further studies may help develop this technique for clinical diagnosis and treatment monitoring in future.

ATR-FTIR spectroscopy and µ-EDXRF spectrometry monitoring of enamel erosion caused by medicaments used in the treatment of respiratory diseases.

Medicaments essential for alleviation of diseases may sometime adversely affect dental health by eroding the enamel, owing to their acidic nature. It is therefore highly desirable to be able to detect these effects quickly and reliably. In this study, we evaluated the erosive capacity of four most commonly prescribed respiratory disease syrup medicaments on enamel using micro-energy-dispersive X-ray fluorescence spectrometry (µ-EDXRF) and attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR). Fifty-five enamel fragments obtained from 30 bovine teeth were treated with artificial saliva (S), acebrofilin hydrochloride (AC), ambroxol hydrochloride (AM), bromhexine hydrochloride (BR), and salbutamol sulfate (SS); by immersing in 3 mL of respective solutions for 1 min, three times a day at intervals of 1 hr, for 5 days. µ-EDXRF analysis of enamel surface did not reveal significant erosion caused by the medications. However, ATR-FTIR showed a detectable shift in the phosphate (PO4 ) antisymmetric stretching mode (ν3 ) at ∼985 cm-1 for AM, BR, and SS, indicating erosion. Multivariate statistical analysis showed that AC, AM, SS, and BR could be classified with 70%, 80%, 100%, and 100% efficiency from S (control), further highlighting the ability of ATR-FTIR to identify degree of erosion. This suggests ATR-FTIR may be used to rapidly and nondestructively investigate erosive effects of medicaments.

Raman spectral post-processing for oral tissue discrimination - a step for an automatized diagnostic system.

Most oral injuries are diagnosed by histopathological analysis of a biopsy, which is an invasive procedure and does not give immediate results. On the other hand, Raman spectroscopy is a real time and minimally invasive analytical tool with potential for the diagnosis of diseases. The potential for diagnostics can be improved by data post-processing. Hence, this study aims to evaluate the performance of preprocessing steps and multivariate analysis methods for the classification of normal tissues and pathological oral lesion spectra. A total of 80 spectra acquired from normal and abnormal tissues using optical fiber Raman-based spectroscopy (OFRS) were subjected to PCA preprocessing in the z-scored data set, and the KNN (K-nearest neighbors), J48 (unpruned C4.5 decision tree), RBF (radial basis function), RF (random forest), and MLP (multilayer perceptron) classifiers at WEKA software (Waikato environment for knowledge analysis), after area normalization or maximum intensity normalization. Our results suggest the best classification was achieved by using maximum intensity normalization followed by MLP. Based on these results, software for automated analysis can be generated and validated using larger data sets. This would aid quick comprehension of spectroscopic data and easy diagnosis by medical practitioners in clinical settings.

Fast monitoring of tooth erosion caused by medicaments used in the treatment of respiratory diseases by ATR-FTIR and µ-EDXRF analysis.

The present study aimed to evaluate the erosive potential of four most commonly prescribed syrup medicaments for respiratory diseases. Attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopy combined with multivariate statistical analysis and micro-energy-dispersive X-ray fluorescence spectrometry (µ-EDXRF) mapping was performed. Fifty-five root dentin fragments obtained from the buccal surface of 30 bovine teeth were prepared and divided into five experimental groups (n = 10): control-artificial saliva (S), acebrofilin hydrochloride (AC), ambroxol hydrochloride (AM), bromhexine hydrochloride (BR), and salbutamol sulfate (SS). The S group was stored only in artificial saliva and the other groups were treated with the medicaments (immersed for 1 min in 3 mL of the medication, three times daily, with 1-h intervals between the immersion cycles, during 5 days, 15 immersion cycles). There were a significant decrease in the Ca and P weight percentages (wt%) for dentin after medication treatments, except for AC (p > 0.05). Mineral content of dentin showed a clear gradation with increasing Ca and P wt% reduction in the order S < AC < AM < BR < SS. SS resulted in a significant increase in Ca/P ratio when compared to the control (p < 0.001). ATR-FTIR combined with multivariate, statistical analysis can quickly and reliably indicate extent of dentin erosion. Considering syrups with high-erosive potential should always follow with proper oral hygiene practices or search for an alternative medications void of such detrimental effects. Regular and prolonged use of these medicaments might bear the risk of causing erosion.