LAMSA.A - Ανάλυση φωνής για μελέτη της προόδου νευρολογικών ασθενειών

LAMSA.A, the speech analysis software could be applied to several diseases towards the early diagnosis as well as the monitoring of the disease progress.

Introduction

The laryngeal mechanism and cognition is subjected to highly complex, extensive neural control; therefore, disorders of the nervous system have effects on the voice. In neurological disorders, dysphonic symptoms are often among the earliest symptoms developed [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[i]].

Dementia

Dementia is a heart-breaking condition that cannot be cured today, but if detected early enough its progress can be slowed. Yet traditional methods of diagnosis and treatment, based on limited professional medical observation, will become increasingly inadequate as the world’s elderly population explodes in the coming decades. In 1995, 30 countries had elderly populations of at least 2 million; by 2030 it’s expected that more than 60 countries will reach this level [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[ii]].Many patients with dementia are placed in nursing homes to receive the round-the-clock monitoring and care they require, which can be stressful and financially challenging for both patients and their families. Diagnosis and treatment also requires doctors and caregivers to observe the patient, but these observations are often limited to the behaviour shown during short visits, making it difficult to accurately diagnose the patient and provide the right type of treatment.

Part of the Dementia condition is the Alzheimer disease. Eight cognitive domains are most often damaged in Alzheimer Disease (AD) [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[iii]], [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[iv]]: memory, language, perception, attention, constructional skills, counselling skills, problem solving, and functional capabilities. The clinical diagnosis is usually based on: Tests of memory and other cognitive functions, behavioural changes analysis; Neuroimaging (CT, SPECT, PET), and the absence of other causes by other medical tests. The greater the number of tests used in the detection, the higher the reliability of the diagnosis. Non-invasive Intelligent Techniques of diagnosis may become valuable tools for early detection of dementia and can be used by non-technologists in the habitual environments of the patient without altering or blocking their abilities. Automatic Spontaneous Speech Analysis is one of them.

After the loss of memory, one of the major problems of AD is the language [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[v]]. The loss of ability to express with language will affect two types or two aspects: difficulty to speak and difficulty to understand others, which difficult the natural communication process with the environment. The problems that the patients have for communicating according to the stage of the disease and how it can help would be [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[vi]]: (i) First Stage: Difficulty for finding the right word in the spontaneous speech. Often it is not detected, (ii) Second Phase: impoverishment of language and vocabulary for everyday use and (iii) Third stage: Answers sometimes are very limited and with very few words.

Multiple Sclerosis

Multiple sclerosis (MS) is a neurological disease with a diverse spectrum of clinical manifestations, among them spastic and ataxic dysarthria are the most commonly observed speech problems [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[vii]].Dysarthria is a motor speech disorder that may be associated with alterations in the regulation of phonation. The phonatory symptoms in MS are often masked by the pronounced scanning speech and abnormal prosodic patterns while speaking. However, the speech is a highly sensitive neurophysiological function, and a detailed study of vocal measures may reflect the status of nervous centres and pathways. Therefore, the study of the speech in carriers of neurological disorders aims at identifying relationships among pathophysiological mechanisms, lesion topography, and distinct phonatory alterations, for example, dysphonic symptoms There are a number of reports about dysarthria in MS patients, but objective speech analyses were used only in a limited number of studies [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[viii]], [http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);">[ix]].

Project Vision

The current efforts could lead to biased results, since the user could pay more attention if he/she feels that is monitored by software. Thus, the proposed work goes beyond state of the art and aims to develop a real-time disease diagnosis and monitoring system, based on speech analysis that tracks speech during everyday activities and during various circumstances.

Along with these characteristic tremors, voice changes are common, such as whispering, breathiness, and a shift to higher tones. In fact, the voice can be weakened by as much as 10 decibels compared to an average speaker. Though these changes can go unnoticed by a person with a disease, relatives and friends more often pick up on it, but a computer algorithm that detects nuances of speech and subtle changes could detect aberrations with much greater frequency.

The technology can scale easily as patients can do the tests themselves in minutes and are as cheap as making a local phone call. This could not only help screen people for early stages of the disease, but it could also allow doctors to track the disease progression in patients and therapists to monitor the effectiveness of voice therapies without patients having to come into the clinic. This could save valuable resources and allow healthcare workers to have more frequent check-ups on patient health remotely.

The proposed work aims to develop a real-time computerized monitoring system, combining a mobile speech recording system with sophisticated speech analysis software and intelligent decision support system towards the effective and early diagnosis as well as the accurate monitoring of disease progress.

 

The architecture of the system will be divided into three tiers:

  • Bottom Tier: Contains the data entry services, i.e. the Data repository that will be developed and populated or/and the Open Access Databases in order to enhance the dataset with more freely available data. In addition, external recording devices (i.e. laptops) are being used in order to record speech and store the information to the local Data Repository.
  • Middle Tier: This tier contains the modules for data analysis. Connecting through secure web-services with the Data Entry Tier, it contains the Speech Analysis software and Decision Support System (DSS). The speech analysis software that will be developed aims to analyse the recorded speech and feeds the DSS with the extracted information. Using intelligent techniques and machine learning algorithms, the DSS analyses information and aims to identify the progress of the disease.
  • Top Tier: Finally, the top tier of the system contains the end user clinician’s application. It is fed by the knowledge extracted by the DSS and the information generated from the Data Analysis tier and displays to the final user (i.e. clinician) the results from the speech analysis process and the Decision Support.

Project Impact

The proposed work aims to develop software to help with patient treatment, so that drug dosage and timing could be optimized. Additionally, clinical trials could benefit from classification methods more accurate than current methods that may fail to detect some with neurological diseases, mentioned before. With a database of recordings available for analysis, more sophisticated algorithms could also be developed that may lead to a scoring system for disease progression based on voice alone.

The potential of this technology goes beyond Parkinson’s disease as voice changes can be caused by other neurological diseases, such stroke, multiple sclerosis, or Lou Gehrig’s disease (ALS), as well as cancer that affects the throat (larynx, esophageal, neck, and even lung cancer). Voice changes also occur with viral and bacterial infection, like the common cold or flu, and are one of the characteristics of heavy smoking. So if similar types of voice recording pools could be collected and analysed with these algorithms, the potential to detect diseases and monitor their progression could be developed.

We're not intending this to be a replacement for clinical experts, rather, it can very cheaply help identify people who might be at high risk of having the disease and for those with the disease, it can augment treatment decisions by providing data about how symptoms are changing in-between check-ups with the neurologist.

This could enable some radical breakthroughs, because voice-based tests are as accurate as clinical tests, but additionally, they can be administered remotely, and patients can do the tests themselves. Also, they are high speed (take less than 30 seconds), and are ultra-low cost (they don't involve expert staff time). So, they are massively scalable. We see the following as having the most impact:

  1. Reduce logistical difficulties in routine practice - no need to visit the clinic for check-ups.
  2. High-frequency monitoring for individualized treatment decisions. With this data, we can optimize drug timing and dosage for maximum effect.
  3. Cost-effective mass recruitment for treatment trials. Recruiting very large numbers into trials for new treatments will speed up the search for a cure.
  4. Population-scale screening programs. Searching for early 'biomarkers' could find the signs of the disease before the damage done are irreparable.

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[i]] Sataloff RT, Mandel S, Rosen DC. Neurologic disorders affecting the voice in performance. In: Sataloff RT, ed. The Professional Voice: The Science and Art of Clinical Care. 2nd ed. San Diego, CA: Singular Publishing Group Inc; 1997:479–498

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[ii]] Panel on a Research Agenda and New Data for an Aging World, Committee on Population, Committee on National Statistics, National Research Council, Preparing for an Aging World: The Case for Cross-National Research (2001).

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[iii]] Morris JC, The Clinical Dementia Rating (CDR): current version and scoring rules. Neurology, 1993. 43: p. 2412b-2414b.

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[iv]] American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental disorders, 4th Edition Text Revision. Washington DC

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[v]] Karmele López-de-Ipiña, Jesús B. Alonso, Nora Barroso, Marcos Faundez-Zanuy, Miriam Ecay, Jordi Solé-Casals, Carlos M. Travieso, Ainara Estanga, Aitzol Ezeiza, New Approaches for Alzheimer’s Disease Diagnosis Based on Automatic Spontaneous Speech Analysis and Emotional Temperature, Ambient Assisted Living and Home Care Lecture Notes in Computer Science Volume 7657, 2012, pp 407-414

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[vi]] Karmele López-de-Ipiña, Jesus-Bernardino Alonso, Carlos Manuel Travieso, Jordi Solé-Casals, Harkaitz Egiraun, Marcos Faundez-Zanuy, Aitzol Ezeiza, Nora Barroso, Miriam Ecay-Torres, Pablo Martinez-Lage and Unai Martinez de Lizardui, On the Selection of Non-Invasive Methods Based on Speech Analysis Oriented to Automatic Alzheimer Disease Diagnosis, Sensors 2013, 13, 6730-6745.

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[vii]] Hartelius L, Runmarker B, Andersen O, et al. Temporal speech characteristics of individuals with multiple sclerosis and ataxic dysarthria: ‘scanning speech’ revisited. Folia Phoniatr Logop. 2000;52:228–238

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[viii]] Feijo AV, Parente MA, Behlau M, Haussen S, et al. Acoustic analysis of voice in multiple sclerosis patients. J Voice. 2004;18:341–347

http://www.lime-technology.gr/components/com_jce/editor/tiny_mce/plugins/anchor/img/anchor.gif);"> [[ix]] Abdul Latif Hamdan, Sahar Farhat, Rami Saadeh, Iyad El-Dahouk, Abla Sibai, and Bassem Yamout, Voice-Related Quality of Life in Patients withMultiple Sclerosis, Hindawi Publishing Corporation Autoimmune Diseases Volume 2012, Article ID 143813.