Evaluation of Vertical Level Differences Between Left and Right Vocal Folds Using Artificial Intelligence System in Excised Canine Larynx

OBJECTIVES

This study aimed to establish an artificial intelligence (AI) system to classify vertical level differences between vocal folds during vocalization and to evaluate the accuracy of the classification.

METHODS

We designed models with different depths between the right and left vocal folds using an excised canine larynx. Video files for the data set were obtained using a high-speed camera system and a color complementary metal oxide semiconductor camera with global shutter. The data sets were divided into training, validation, and testing. We used 20,000 images for building the model and 8000 images for testing. To perform deep learning multiclass classification and to estimate the vertical level difference, we introduced DenseNet121-ConvLSTM.

RESULTS

The model was trained several times using different numbers of epochs. We achieved the most optimal results at 100 epochs, and the batch size used during training was 16. The proposed DenseNet121-ConvLSTM model achieved classification accuracies of 99.5% and 88.0% for training and testing, respectively. After verification using an external data set, the overall accuracy, precision, recall, and f1-score were 90.8%, 91.6%, 90.9%, and 91.2%, respectively.

CONCLUSIONS

The newly developed AI system may be an easy and accurate method for classifying superior and inferior vertical level differences between vocal folds. Thus, this AI system can be applied and may help in the assessment of vertical level differences in patients with unilateral vocal fold paralysis.

Validation and development of Voice Handicap Index-Throat (VHI-T) for Koreans

INTRODUCTION

This study aimed to develop, validate, and analyze the reliability of the Korean version of the Voice Handicap Index-Throat (VHI-Tk).

METHODS

This prospective study included 103 patients in the case group with voice problems (18 with functional dysphonia, 44 with mass in the larynx, 18 with neurological voice disorder, 23 with throat problems) and 27 in the control group without voice problems. All participants completed these questionnaires at their initial visit: the Korean version of the Voice Handicap Index (K-VHI), VHI-Tk, and the Korean version of the Voice Symptom Scale (K-VoiSS). Case group patients in the case group re-completed the VHI-Tk questionnaire to assess test-retest reliability. Finally, a one-way analysis of variance was implemented to assess differences in VHI-Tk scores among the four diagnosis types in the case group.

RESULTS

The VHI-Tk scores in the case group were significantly higher than in the control group. The VHI-Tk was significantly correlated with the subscales of K-VHI and K-VoiSS. The VHI-Tk has significant test-retest reliability, and its internal consistency is good to excellent (Cronbach's alpha correlation coefficient range: 0.895-0.901). There was significant difference in the mean VHI-Tk scores according to the four diagnosis types (throat problems group > neurological voice disorder group).

CONCLUSION

We validated the VHI-T questionnaire to measure self-perceived voice and throat problems among Koreans. A large sample size and various diagnosis types are required in future studies to fully validate the VHI-T for use in multiple cultures.

A Cepstral Analysis of Pathological Voice Quality in the Korean Population using Praat

OBJECTIVES

This study aimed to investigate the reference values for cepstral peak prominence (CPP) and smoothed CPP (CPPS) measured using Praat in Korean speakers with the normal, healthy and pathological voice.

METHODS

A total of 4,524 Korean participants with vocally healthy (n = 410) and dysphonic voices (n = 4,114) participated in this study. The speech task consisted of a sustained vowel /a/ and a sentence reading the Korean passage "Walk". CPP and CPPS values were quickly and automatically measured in sustained vowel and continuous speech tasks using Praat script. Furthermore, three veteran speech language pathologists (SLPs) scored the severity of dysphonia using the GRBAS scale (grade, roughness, breathiness, asthenia, strain) and Consensus Auditory Perceptual Evaluation of Voice (CAPE-V).

RESULTS

Three SLPs showed high inter- and intra-rater reliabilities (IRR) in auditory-perceptual (A-P) evaluation. Significant differences were confirmed in CPP and CPPS between the normally healthy and pathological voice groups for both voice tasks (P < 0.01). The measured values of CPP and CPPS varied depending on the laryngeal pathology. In the receiver operating characteristic (ROC) curve analysis, the CPP_Vowel (CPP_V), CPPS_V, CPP_Sentence (CPP_S), and CPPS_S cut-off values were <21.5, <12.0, <19.7, and <10.1, respectively. Through ROC curve analysis, it was confirmed that CPP and CPPS had excellent diagnostic accuracy in distinguishing disordered voice (area under the ROC: 0.951-0.966).

CONCLUSION

We investigated the reference values for CPP and CPPS measured with Praat for Korean speakers and confirmed that cepstral analysis is a promising tool for differentiating pathological voice.

Diagnosis of Early Glottic Cancer Using Laryngeal Image and Voice Based on Ensemble Learning of Convolutional Neural Network Classifiers

OBJECTIVES

The purpose of study is to improve the classification accuracy by comparing the results obtained by applying decision tree ensemble learning, which is one of the methods to increase the classification accuracy for a relatively small dataset, with the results obtained by the convolutional neural network (CNN) algorithm for the diagnosis of glottal cancer.

METHODS

Pusan National University Hospital (PNUH) dataset were used to establish classifiers and Pusan National University Yangsan Hospital (PNUYH) dataset were used to verify the classifier's performance in the generated model. For the diagnosis of glottic cancer, deep learning-based CNN models were established and classified using laryngeal image and voice data. Classification accuracy was obtained by performing decision tree ensemble learning using probability through CNN classification algorithm. In this process, the classification and regression tree (CART) method was used. Then, we compared the classification accuracy of decision tree ensemble learning with CNN individual classifiers by fusing the laryngeal image with the voice decision tree classifier.

RESULTS

We obtained classification accuracy of 81.03 % and 99.18 % in the established laryngeal image and voice classification models using PNUH training dataset, respectively. However, the classification accuracy of CNN classifiers decreased to 73.88 % in voice and 68.92 % in laryngeal image when using an external dataset of PNUYH. To solve this problem, decision tree ensemble learning of laryngeal image and voice was used, and the classification accuracy was improved by integrating data of laryngeal image and voice of the same person. The classification accuracy was 87.88 % and 89.06 % for the individualized laryngeal image and voice decision tree model respectively, and the fusion of the laryngeal image and voice decision tree results represented a classification accuracy of 95.31 %.

CONCLUSION

The results of our study suggest that decision tree ensemble learning aimed at training multiple classifiers is useful to obtain an increased classification accuracy despite a small dataset. Although a large data amount is essential for AI analysis, when an integrated approach is taken by combining various input data high diagnostic classification accuracy can be expected.

DNA transfection using linear poly(ethylenimine) prepared by controlled acid hydrolysis of poly(2-ethyl-2-oxazoline)

A series of linear poly(ethylenimine) (L-PEI) containing varying amounts of cationic charge density in its backbone was produced by controlled hydrolysis of poly(2-ethyl-2-oxazoline) (PEtOz) for using as a nonviral DNA transfection agent. The effects of cationic charge density and molecular weight of the L-PEI on the cytotoxicity and transfection efficiency were studied. The efficiency of transfection was monitored by using a luciferase reporter gene system. Gel retardation assay and dynamic light scattering (DLS) showed that the condensation capacity of L-PEI was suitable for transfection. Highly compacted L-PEI/DNA complex ( approximately 150 nm) was obtained with a surface charge value of around +28.4 mV. Cell cytotoxicity was affected to a great extent by the hydrolysis percent of L-PEI as well as by the molecular weight. Transfection efficiency of luciferase plasmid DNA against NIH 3T3 fibroblast was largely dependent upon the hydrolysis percent (charge density) in the polymer backbone and the molecular weight of the L-PEI, but independent of the total amount of cationic charges used for DNA condensation. L-PEI with a hydrolysis percent of 88.0% exhibited comparable transfection efficiency to that of commonly used branched PEI.

Normative measurements of Korean orbital structures revealed by computerized tomography

PURPOSE

The purposes of this study were to establish criteria for the diameters of normal extraocular muscles, the width of the optic nerve, and the globe position as revealed by CT, and to investigate the effects of age and sex on these parameters in the Korean population.

SUBJECTS AND METHODS

Diameters of extraocular muscles (medial, lateral, superior complex, and inferior rectus), distance from the interzygomatic line to the posterior margin of the globe, width of the optic nerve-sheath complex, and length of the interzygomatic line were calculated for 214 patients on axial and direct coronal CT images, and the effects of age and sex were analyzed.

RESULTS

Normal range for the diameters as given by mean+/-2SDs of extraocular muscles were medial rectus, 2.2--5.4 mm; lateral rectus, 2.1--4.9 mm; inferior rectus, 2.5--5.7 mm; and superior complex group, 2.6--6.4 mm. The mean optic nerve-sheath complex was 4.1 mm and the values ranged from 2.9 mm to 5.3 mm. The mean length of the interzygomatic line was 103.8 mm. The normal posterior position of the globe as here designed was 11.2 mm behind the interzygomatic line (range, 6.4--15.3 mm). All parameters under study were a little larger in males than females. Statistically, however, there was no significant difference between males' and females' results, nor between the various age groups (p>0.05).

CONCLUSION

Our CT results may help clinical ophthalmologists to accurately assess enlargement of the extraocular muscles and the optic nerve as well as aiding in the evaluation of exophthalmos.

Poly(DMAEMA-NVP)-b-PEG-galactose as gene delivery vector for hepatocytes

A block copolymer composed of cationic polymer and poly(ethylene glycol) (PEG) was used as a DNA carrier. Poly(2-(dimethylamino)ethyl methacrylate (DMAEMA)-co-N-vinyl-2-pyrrolidone (NVP)) having a terminal carboxylic group was synthesized by free radical polymerization using an initiator, 4,4'-azobis(4-cyanovaleric acid). The terminal carboxylic acid was activated by N-hydroxysuccinimide (NHS) with dicyclohexylcarbodiimide (DCC) and then conjugated with PEG-bis(amine). For specific gene targeting to asialoglycoprotein receptor of hepatocytes, a galactose moiety was incorporated into the PEG terminal end of poly(DMAEMA-NVP)-b-PEG by reductive coupling using lactose and sodium cyanoborohydride. RSV luciferase plasmid was used as a reporter gene, and in vitro gene transfection efficiency was measured in HepG2 human hepatocarcinoma cells. Poly(DMAEMA-NVP)-b-PEG-galactose/DNA complexes formed at 0.5-2 polymer/plasmid weight ratio had compacted structures around 200 nm particle size and exhibited slightly negative surface charge. These complexes were coated with a cationic, pH sensitive, endosomolytic peptide, KALA, to generate positively charged poly(DMAEMA-NVP)-b-PEG-galactose/DNA/KALA complex particles. In the presence of serum proteins, both the PEG block and the galactose moiety of poly(DMAEMA-NVP)-b-PEG-galactose greatly enhanced the gene transfection efficiency, which was very close to that of Lipofectamine plus. Irrespective of the presence of serum proteins, as the KALA/DNA weight ratio increased, the transfection efficiency of poly(DMAEMA-NVP)-b-PEG-galactose was enhanced due to the pH dependent endosomal disruptive property of KALA. This study demonstrates that sufficient transfection efficiency as high as that of commercial agent could be attained by judicious formulation of molecular engineered poly(DMAEMA-NVP)-b-PEG-galactose in combination with an endosomolytic peptide, KALA.

Nasal cycle in patients with septal deviation: evaluation by acoustic rhinometry

The nasal cycle in patients with septal deviation was studied by acoustic rhinometric techniques. This study included 24 patients with anteriorly located septal deviations (mean age = 23.5), and 26 normal controls (mean age = 24.7). Data of MCA (minimum cross-sectional area) and NV (nasal volume), collected in 20-minute intervals, were plotted for each subject during 8 hours. Twenty of 24 patients (83%) with septal deviation and 20 of 26 normal subjects (77%) showed at least one complete cycle. Duration of the nasal cycle, which ranged from 100 minutes to 400 minutes, had no statistical difference between the septal deviation group (mean duration of 216 minutes) and the normal control group (mean duration of 227 minutes). The degrees of variation of MCA and NV, defined as Degree of Variation of MCA (%) = 100 (MCAmax - MCAmin)/MCAmax, Degree of Variation of NV (%) = 100 (NVmax - NVmin)/NVmax, which represent the percent change of MCA and NV throughout the study, showed no difference between the wide side and the narrow side, or between the septal deviation group and the normal control group. These findings suggest that the nasal cycle is relatively independent of peripheral anatomic factors for its generation. However, the amplitude of changes of MCA was greater in the wide side, and the sum of both MCAs tended to fluctuate in accordance with the fluctuation of MCA of the wide side. Thus, the nasal cycle seemed to be affected by septal deviation.

Effects of the acoustic noise of the gradient systems on fMRI: a study on auditory, motor, and visual cortices

MR acoustic, or sound, noise due to gradient pulsing has been one of the problems in MRI, both in patient scanning as well as in many areas of psychiatric and neuroscience research, such as brain fMRI. Especially in brain fMRI, sound noise is one of the serious noise sources that obscures the small signals obtainable from the subtle changes occurring in oxygenation status in the cortex and blood capillaries. Therefore, we have studied the effects of acoustic, or sound, noise arising in fMR imaging of the auditory, motor, and visual cortices. The results show that the effects of acoustic noise on motor and visual responses are opposite. That is, for motor activity, there is an increased total motor activation, whereas for visual stimulation, the corresponding (visual) cortical activity is diminished substantially when the subject is exposed to a loud acoustic sound. Although the current observations are preliminary and require more experimental confirmation, it seems that the observed acoustic-noise effects on brain functions, such as in the motor and visual cortices, are new observations and could have significant consequences in data observation and interpretation in future fMRI studies.

NsiI and ScaI restriction fragment length polymorphisms at the atrial natriuretic peptides (ANP) gene locus

The authors report on two new restriction fragment length polymorphisms at the human atrial natriuretic peptide gene locus, detected in three families with restriction endonucleases ScaI and NsiI.

Human apolipoprotein CI (ApoC1) gene locus: BglI dimorphic site

Images

BanI dimorphic site in the third intron of the human apolipoprotein AI gene (Apo A1)

Images

DraI RFLP in the human apolipoprotein AI-CIII-AIV gene complex

Images